CN114519302A - Road traffic situation simulation method based on digital twin - Google Patents
Road traffic situation simulation method based on digital twin Download PDFInfo
- Publication number
- CN114519302A CN114519302A CN202210099127.8A CN202210099127A CN114519302A CN 114519302 A CN114519302 A CN 114519302A CN 202210099127 A CN202210099127 A CN 202210099127A CN 114519302 A CN114519302 A CN 114519302A
- Authority
- CN
- China
- Prior art keywords
- model
- twin
- road
- target object
- simulation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000004088 simulation Methods 0.000 title claims abstract description 99
- 238000000034 method Methods 0.000 title claims abstract description 75
- 238000011156 evaluation Methods 0.000 claims abstract description 30
- 238000007637 random forest analysis Methods 0.000 claims abstract description 22
- 238000012549 training Methods 0.000 claims abstract description 11
- 230000003042 antagnostic effect Effects 0.000 claims abstract description 10
- 239000003795 chemical substances by application Substances 0.000 claims description 56
- 238000004458 analytical method Methods 0.000 claims description 40
- 238000004364 calculation method Methods 0.000 claims description 40
- 238000003066 decision tree Methods 0.000 claims description 34
- 230000006870 function Effects 0.000 claims description 26
- 239000011159 matrix material Substances 0.000 claims description 18
- 238000004891 communication Methods 0.000 claims description 17
- 238000000605 extraction Methods 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 16
- 230000009471 action Effects 0.000 claims description 13
- 238000007781 pre-processing Methods 0.000 claims description 13
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 230000003993 interaction Effects 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 230000000007 visual effect Effects 0.000 claims description 6
- 230000009467 reduction Effects 0.000 claims description 5
- 206010039203 Road traffic accident Diseases 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 4
- 230000002776 aggregation Effects 0.000 claims description 3
- 238000004220 aggregation Methods 0.000 claims description 3
- 230000004927 fusion Effects 0.000 claims description 3
- 238000012552 review Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000013210 evaluation model Methods 0.000 abstract description 2
- 238000012800 visualization Methods 0.000 description 13
- 238000010586 diagram Methods 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000013079 data visualisation Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000010205 computational analysis Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000013107 unsupervised machine learning method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a road traffic situation simulation method based on digital twins, which is used for evaluating traffic situations of different road sections. The method designs a corresponding free forest model, condenses high-influence factors in a clustering mode, outputs important attributes influencing traffic situation, designs a Bayesian network model to form a probability judgment module, and establishes a traffic simulation model in a multi-agent environment. An advanced agent framework based on analog is built through a threshold value method to obtain a target object simulation model, and the target object simulation model and a twin simulation path model based on a Tensorflow platform are used for performing antagonistic learning and outputting an evaluation result. The method comprises the steps of communicating a twin target object model with an actual situation through a digital twin information channel, training and deploying a real-time random forest model, and carrying out real-time simulation and real-time update on the traffic situation through the digital twin road traffic situation evaluation model. The invention can be used for effective traffic situation assessment in various traffic scenes.
Description
Technical Field
The invention relates to the technical field of traffic situation simulation, in particular to a road traffic situation simulation method based on digital twins.
Background
Along with the development of cities, the importance of road traffic is more prominent in the development of the society and the economy in China. The increasing scale and complexity of traffic networks lead to more and more serious problems of traffic jam, traffic accidents and the like. Whether the department of transportation function or the traveler has higher and higher requirements on the safety and convenience of transportation, and the requirement provides a challenge for the accuracy of the perception of the traffic situation. Therefore, the road traffic situation simulation is realized as accurately as possible by utilizing advanced technology and detection equipment, the traffic condition of the current road section is grasped in real time, the travel efficiency is improved, the traffic safety is enhanced, and the method has important theoretical and practical research significance in the aspects of guiding information technology application, promoting urban comprehensive competitiveness, realizing sustainable development and the like.
With the continuous development of detection equipment and data transmission functions, large-scale multi-dimensional and real-time traffic data can be rapidly collected. In the prior art, most of models predict short-term traffic situations by analyzing time sequence characteristics of traffic data by using an artificial intelligence method such as a neural network, but most of models do not reflect the influence of uncertain factors under the time sequence characteristics on the traffic situations, and the accuracy and the real-time updating rate are required to be improved.
Disclosure of Invention
The purpose of the invention is as follows: in view of the above problems, the present invention provides a road traffic situation simulation method based on digital twin, which can simulate the traffic situation, perform real-time simulation and real-time update, and efficiently and accurately realize the prediction of the traffic retardation of the predicted target road section.
The technical scheme is as follows: a road traffic situation simulation method based on digital twins comprises the following steps:
s001: the calculation analysis module acquires the information of the target object and the information of the space where the target object is located through the information extraction module;
s002: reading and preprocessing twin target object characteristic information, modeling by a calculation analysis module through a digital twin-based modeling method, completing model self-optimization, selecting an optimal model for simulation, and generating a road situation evaluation result;
s003: sending the road situation evaluation result to a preset receiving end through a communication module;
s004: and visually converting the road situation evaluation result, and generating visual video data on a visual module in a high frame rate video mode.
Further, in S002, the model created by the digital twin-based modeling method includes a twin target object feature model, a twin target object model, and a twin simulation path model.
Further, the twin target object feature model is built by adopting a random forest modeling method, and comprises the following steps,
s101, a calculation analysis module acquires road characteristic data after noise reduction and fusion of traffic road conditions through an information extraction module;
s102: the calculation analysis module reads and preprocesses the noise-reduced and fused road characteristic data of the traffic road condition, and carries out modeling analysis on the preprocessed noise-reduced and fused road characteristic data through a random forest model to generate an importance sequencing result;
s103, the calculation analysis module provides road characteristic data with the importance lower than 5%, and outputs a twin target object characteristic result;
and S104, storing the obtained twin target object characteristic result into a data set in a matrix form to form a twin target object characteristic model.
Further, in S102, the modeling analysis of the random forest model includes the following steps,
s11: averagely dividing the total tree number M in the forest into 4 parts, distributing the 4 parts to each process for sharing, creating M/4 decision trees by each process, and returning the created M/4 decision trees to the main process in a list form;
s12: after obtaining the decision tree lists of 4 sub-processes, the main process integrates the 4 sub-lists into a decision tree list L with the length of M;
s13: generating a training set for each decision tree, and repeatedly and randomly extracting n samples from an original training sample set by using a bootstrapping aggregation method in bagging to serve as a group of training sets;
s14: constructing a single decision tree, randomly extracting m features from the features of each sample, dividing the m features into a sample set B in a feature-based mode to calculate a Keyny index, and setting the sample set as a sample set D1 equal to a specified feature and a sample set B2 not equal to the given feature, then:
wherein A is the current characteristic, then finding out the division A with the minimum Gini index from all Gini (B, Ai), using the characteristic A as the splitting characteristic, and repeating the step to obtain a decision tree;
s15: each process repeatedly constructs M/3 decision trees according to the steps S13 and S14, the decision trees are integrated to form a random forest, and the last leaf node of each decision tree is the feature with the largest influence on network security to form a feature set A1 and A2 … … Am;
s16: judging the effectiveness of the model according to the accuracy and selecting the optimal parameters:
wherein TP is true positive, which indicates that the actual category is positive, and the algorithm output category is also positive; FP is false positive, which means that the actual category is negative, and the sample output category is positive; TN is true negative, which means that the actual category is negative, and the algorithm output category is also negative; FN is false negative, which indicates that the actual category is positive, and the algorithm output category is negative; the hyper-parameter θ is set to: max _ features is 0.2 and decision tree number tree is 18.
Furthermore, the twin target object model is established by adopting a method for modifying the original Agent, and comprises the following steps,
s31: an initialization stage: the vehicle physical entity is initialized to Agent, and the attribute of the Agent is defined as:
wherein id represents the unique identification number of the Agent;
representing a velocity vector of the Agent; pos (t) represents the location of Agent; h represents the length of the Agent; r represents the width of the Agent; p represents the average speed of the Agent; t represents a time step;
s32: and (3) an interaction stage: each Agent carries out information interaction with other agents and the environment where the agents are located, and the current position is recorded in a coordinate mode;
s33: an action stage: each Agent moves according to the driving of the social force model, and continuously updates the position information until the action stops;
s34: and recording and returning the speed change condition of the action on the road section, and judging and returning the road traffic slow-delay condition according to the occupation ratio of the agent action low-speed time section in the running time section.
Preferably, the twin simulation path model is built and trained using a bayesian network model, comprising the steps of,
s21: feature sample set { A) obtained by random forest model learning simulation1，A2……AmCombining the occurrence probability with the corresponding occurrence time period and a sample label to form a feature set sample T with an occurrence probability value, recording the occurrence probability as prior probability, updating the feature sample, correcting all the prior probabilities by using a Bayes formula, and solving the posterior probability;
s22: taking parameters related to the road section traffic situation as nodes, wherein the nodes comprise observable nodes and hidden nodes; the observable nodes comprise the average travel time of the road sections and the relative density of the road sections, which can cause traffic, and the hidden nodes comprise the traffic situation of the road sections and the historical probability of traffic accidents of the road sections;
s23: and calculating the probability of traffic delay of the corresponding road section by adopting a Bayesian formula, wherein the Bayesian formula is as follows:
where P (A | B) is the likelihood of A occurring in the case of B occurring, A1, … …, Ai are complete event groups, P (A) is the prior probability or marginal probability of A, P (A | B) is the conditional probability of A after B is known to occur, P (B | A) is the conditional probability of B after A is known to occur, and P (B) is the prior probability or marginal probability of B.
S24: and forming a Bayesian network for predicting the traffic delay condition of the road, and outputting a result.
Preferably, the results output from the Bayesian network of S24 are used as the results of the simulation of the object by the twin target object model for the comprehensive review of the antagonistic learning, the antagonistic learning model is established,
the result output by the twin target object model is reviewed by adopting the result output by the twin simulation path model, and the result is evaluated according to a comprehensive evaluation method, which comprises the following steps,
s41: a loss function gloss is generated, and the formula is:
loss＝ganWeight×ganLoss+modelWeight×modelLoss；
wherein the ganWeight is the weight of the loss-resisting function ganLoss, and the value is 0-30; the modelWeight is the weight of a loss function modelLoss of a difference value between a twin target object model output result and a twin simulation path model output result, and the value of the modelWeight is 0-50;
modelLoss represents the difference between the output result of the twin target object model and the output result of the twin simulation path model, and the calculation formula is as follows:
wherein, RPijRepresenting the twin target object model output result, QPijRepresenting the output result of the twin simulation path model, wherein the two results are represented in a two-dimensional matrix mode;
s42: generating ganLoss as a function of the countermeasure loss, wherein the formula is as follows:
wherein, Pmodel is a probability matrix that the counterlearning model considers that the result output by the twin target object model is completely consistent with the result output by the twin simulation path model when Pmodel is counterlearning, the matrix is a two-dimensional matrix and accommodates the two-dimensional matrixPmodel values of different roads in different time periods; the EPCG is a dimension stability factor when the gradient is reduced, and the EPCG adopts a pseudorandom function in the simulation of [10 ]-4,10-11]Taking a value in the interval in a pseudo-random mode;
s43: and (3) adopting a discrimination formula to discriminate the random function, wherein the smaller the loss function is, the more accurate the simulation result is, wherein the discrimination formula is as follows:
wherein,
is the desired probability, i.e. the probability matrix whose actual situation is consistent with the result output by the twin target object model.
Has the advantages that: compared with the prior art, the invention has the advantages that:
(1) the complexity and systematicness of the traffic situation are comprehensively analyzed, a modeling method based on digital twins is adopted for carrying out traffic situation simulation evaluation, a traffic situation evaluation modeling theory based on digital twins and a modeling method are provided, the traffic situation of different road sections in different time periods is predicted through simulation, scientific basis is provided for traffic prediction, traffic control and traffic planning, and the method has obvious guiding significance;
(2) establishing a digital twin traffic situation evaluation simulation model: establishing a random forest model, and obtaining twin target object characteristics by an unsupervised machine learning method; and establishing a twin simulation path model and a twin target object simulation model, optimizing the models through a countermeasure learning method based on the actual traffic situation, and outputting a simulation result.
(3) And performing digital twin modeling in the Tensorflow environment, deepening the application level of the Tensorflow environment, simulating and analyzing the influence of important attributes on the traffic situation, and providing a related suggestion of traffic planning.
(4) Based on the digital twin modeling method, a new method for model establishment and model self-optimization is provided, on the basis of completing the application of the invention, the modeling direction in the traffic simulation field is widened, and the digital twin self-optimization type modeling method is provided.
(5) The digital twin is used as a new artificial intelligence method, on the basis of the complex time relevance of the traffic situation, a Bayesian network is combined, a higher-level situation perception technology is utilized, factors influencing the traffic situation are extracted, and the method has a good effect on solving the uncertain influence in the traffic situation. The method can accurately predict the traffic situation simulation condition of the target road section in real time, continuously updates the result along with the evolution of the situation, and has important significance on the work of traffic situation prediction, control, traffic network planning and the like.
Drawings
FIG. 1 is a schematic diagram of simulation method module signaling;
FIG. 2 is a flow chart of decision tree parallel distribution of a random forest algorithm;
FIG. 3 is a diagram of a Bayesian network based twin path model;
FIG. 4 is a schematic diagram of a modified Agent;
FIG. 5 is a flow chart of core prediction model building.
Detailed Description
The present invention will be further illustrated with reference to the following figures and specific examples, which are to be understood as merely illustrative and not restrictive of the scope of the invention.
A road traffic situation simulation method based on digital twins comprises four modules, as shown in FIG. 1, including an information extraction module 100, a calculation analysis module 200, a communication module 300 and a visualization module 400.
The calculation analysis module 200 comprises a data reading submodule, a data preprocessing submodule, a core model submodule and a data output submodule which are sequentially connected through an API (application program interface), the information extraction module 100 is in signal connection with the data reading submodule, the communication module 300 and the visualization module 400 are respectively in signal connection with the data output submodule, and the visualization module 400 comprises a preset receiving end which is in signal connection with the communication module 300.
The information extraction module 100 is configured to detect a target object and information of a road segment where the target object is located, where the target object is a vehicle driving on a road in the present invention, and the information extraction module 100 sends the acquired data information to the calculation and analysis module 200.
The calculation analysis module 200 is in signal connection with the communication module 300 and the visualization module 400, and is configured to automatically establish a multi-digital twin-based road traffic situation assessment model, provide calculation power for assessment simulation, and send a traffic situation prediction result to the communication module 300 and the visualization module 400. The communication module 300 is configured to transmit information and send a received road condition evaluation result to a preset receiving end. The visualization module 400 is used for visualizing and displaying the road condition evaluation result.
As shown in FIGS. 2-5, the road traffic situation simulation method based on the digital twin comprises the following steps:
s001: the calculation analysis module acquires the information of the target object and the information of the space where the target object is located through the information extraction module;
s002: reading and preprocessing twin target object characteristic information, modeling by a calculation analysis module through a digital twin-based modeling method, completing model self-optimization, selecting an optimal model for simulation, and generating a road situation evaluation result;
s003: sending the road situation evaluation result to a preset receiving end through a communication module;
s004: and visually converting the road situation evaluation result, and generating visual video data on a visual module in a high frame rate video mode.
The computational analysis module 200 is an emulated computer environment integrated in a computer, mainly in python language, in which: the data reading submodule, the data preprocessing submodule, the core model submodule and the data output submodule are four functional partitions in the environment. The core model submodule is a calculation analysis module 200 for storing and operating the digital twin-based road traffic situation assessment model mentioned in the invention.
And the data reading sub-module is used for reading the data information of the target object and the space where the target object is located, which is extracted by the information extraction module. And the data preprocessing submodule is used for preprocessing the target object and the data information of the space where the target object is located.
And the core model submodule is used for automatically establishing a digital twin-based road traffic situation assessment model and simulating the traffic situation through the digital twin-based road traffic situation assessment model to output an assessment result.
The model built by the modeling method based on the digital twin comprises a twin target object characteristic model, a twin target object model and a twin simulation path model.
Carrying out feature merging processing on road feature data of real-time traffic road conditions to obtain important feature attributes of lanes; denoising road section characteristic data in different time periods, removing noise data, and obtaining denoised road section characteristic data in corresponding time periods; carrying out data fusion processing on the road section characteristic data in different time periods after denoising, and fusing the same type data to obtain the fused road characteristic attribute of the corresponding time period; according to the fused road characteristic data, importance identification processing is carried out on the fused road characteristic attributes, specifically, a random forest modeling method is adopted to carry out unsupervised machine learning training, a soft classification method is adopted to carry out the fused road characteristic classification through a trained random forest model, whether the road has a slow-delay condition in the same time period is taken as a judgment label, the slow-delay state in the invention means that the flow rate of vehicles is lower than 10% of the average flow rate, the characteristic attributes with the corresponding road importance degree lower than 5% (too low) are removed, and a twin target object characteristic result is obtained by outputting, wherein the twin target object characteristics comprise the characteristic attributes of a traffic road environment and the characteristic attributes of vehicles moving in the traffic road environment.
The twin target object feature model is built by adopting a random forest modeling method and comprises the following steps,
s101, a calculation analysis module acquires road characteristic data after noise reduction and fusion of traffic road conditions through an information extraction module;
s102: the calculation analysis module reads and preprocesses the noise-reduced and fused road characteristic data of the traffic road condition, and carries out modeling analysis on the preprocessed noise-reduced and fused road characteristic data through a random forest model to generate an importance sequencing result;
s103, the calculation analysis module provides road characteristic data with the importance lower than 5%, and outputs a twin target object characteristic result;
and S104, storing the obtained twin target object characteristic result into a data set in a matrix form to form a twin target object characteristic model.
The modeling analysis of the random forest model comprises the following steps:
s11: averagely dividing the total tree number M in the forest into 4 parts, distributing the 4 parts to each process for sharing, creating M/4 decision trees by each process, and returning the created M/4 decision trees to the main process in a list form;
s12: after obtaining the decision tree lists of 4 sub-processes, the main process integrates the 4 sub-lists into a decision tree list L with the length of M;
s13: generating a training set for each decision tree, and randomly extracting n samples from the original training sample set by using a bootstrapping aggregation method in bagging as a group of training sets;
s14: constructing a single decision tree, randomly extracting m features from the features of each sample, dividing the m features into a sample set B in a feature-based mode to calculate a Keyny index, and setting the sample set as a sample set D1 equal to a specified feature and a sample set B2 not equal to the given feature, then:
wherein A is the current characteristic, then finding out the division A with the minimum Gini index from all Gini (B, Ai), using the characteristic A as the splitting characteristic, and repeating the step to obtain a decision tree;
s15: each process repeatedly constructs M/3 decision trees according to the steps S13 and S14, the decision trees are integrated to form a random forest, and the last leaf node of each decision tree is the feature with the largest influence on network security to form a feature set A1 and A2 … … Am;
s16: judging the effectiveness of the model according to the accuracy and selecting the optimal parameters:
wherein TP (true positive) is true positive, which indicates that the actual category is positive, and the algorithm output category is also positive; FP (false positive) is false positive, which indicates that the actual category is negative and the sample output category is positive; TN (true negative) indicates that the actual category is negative, and the algorithm output category is also negative; FN (false negative) indicates that the actual class is positive and the algorithm output class is negative. The closer the Accuracy (Accuracy) is to 1, the higher the model Accuracy. The hyper-parameter θ is set to: the RFB module with max _ features ═ 0.2 and decision tree number tree ═ 18 is used in the modeling of the present invention.
The twin target object model is established by adopting a method for modifying an original Agent, and the method comprises the following steps of:
s31: an initialization stage: the vehicle physical entity is initialized to be Agent, and the attribute of the Agent after modification is defined as:
wherein id represents the unique identification number of the Agent;
representing a velocity vector of the Agent; pos (t) represents the location of Agent; h represents the length of the Agent; r represents the width of the Agent; p represents the average speed of the Agent; t represents a time step; in the present invention, the shape of the Agent is modified to a rounded rectangle V01 as shown in fig. 4 in order to fit the vehicle shape.
S32: and (3) an interaction stage: each Agent carries out information interaction with other agents and the environment where the agents are located, and the current position is recorded in a coordinate mode;
s33: an action stage: each Agent moves according to the driving of the social force model, and continuously updates the position information until the action stops;
s34: and recording and returning the speed change condition of the action on the road section, and judging and returning the road traffic slow-delay condition according to the occupation ratio of the agent action low-speed time section in the running time section.
Preferably, the social force model of the modified Agent is:
wherein, (x, y) is the centroid coordinate of the Agent, (x0, y0) is the centroid coordinate of other agents, and r is the minimum distance from the centroid coordinate of other agents to the boundary; x and Y are reduction distances in the X direction and the Y direction, and the reduction distances can reduce the Agent into a right circular V02 in simulation calculation.
Maintaining a force for the Agent to pass through a historical average speed for the road segment;
forces brought to other agents, including repulsion, squeezing, friction between agents;
the random change of the Agent motion is described for the behavior fluctuation, including the reversing action.
According to the characteristics of the twin target object, the twin simulation path model simulates and outputs a traffic incident evaluation result based on the Bayesian network; according to the characteristics of the twin target object, the twin target object simulation model simulates and outputs a traffic event evaluation result based on the modified Agent model, and the twin simulation path model is established and trained by adopting a Bayesian network model and comprises the following steps:
s21: feature sample set { A) obtained by random forest model learning simulation1，A2……AmCombining the occurrence probability with the corresponding occurrence time period and a sample label to form a feature set sample T with the occurrence probability value, marking the occurrence probability as prior probability, updating the feature sample, correcting all the prior probabilities by using a Bayesian formula, and solving the posterior probability;
s22: taking parameters related to the road section traffic situation as nodes, wherein the nodes comprise observable nodes and hidden nodes; the observable nodes comprise the average travel time of the road sections and the relative density of the road sections, which can cause traffic, and the hidden nodes comprise the traffic situation of the road sections and the historical probability of traffic accidents of the road sections;
s23: calculating the probability of traffic delay of the corresponding road section by adopting a Bayesian formula, wherein the Bayesian formula is as follows:
where P (A | B) is the likelihood of A occurring in the case of B occurrence A1, … …, Ai is the complete event group, and P (A) is the prior probability or edge probability of A. This is called a priori because it does not take into account any B-aspect factors. P (a | B) is the conditional probability of a after B is known to occur, and is also referred to as a posterior probability of a due to the value derived from B. P (B | a) is the conditional probability of B after a is known to occur, and is also referred to as the a posteriori probability of B due to the value derived from a. P (B) is the prior probability or edge probability of B, also taken as a normalization constant.
S24: a bayesian network for predicting the traffic sluggish condition of the road is formed and the result is output, as shown in fig. 3, which is a schematic diagram of the bayesian network generated by the present invention.
The bayesian network formed in this embodiment has the following nodes: peak hours G1, peak hours ratio G2, weather G3, road length G4, road width G6, vehicle density H1, average vehicle speed H2, road condition K.
Taking the output result of the Bayesian network of S24 as the result of the simulation of the object for the twin target object model to carry out the antagonistic learning comprehensive review, establishing an antagonistic learning model,
the method comprises the following steps of adopting a twin simulation path model output result closest to the actual situation to comment on a twin target object model output result, and evaluating the result according to a comprehensive evaluation method, wherein the comprehensive evaluation method comprises the following steps:
s41: a loss function gloss is generated, and the formula is:
loss＝ganWeight×ganLoss+modelWeight×modelLoss；
wherein the ganWeight is the weight of the loss-resisting function ganLoss, and the value is 0-30; the modelWeight is the weight of a loss function modelLoss of a difference value between a twin target object model output result and a twin simulation path model output result, and the value of the modelWeight is 0-50;
modelLoss represents the difference between the output result of the twin target object model and the output result of the twin simulation path model, and the calculation formula is as follows:
wherein, RPijRepresenting the twin target object model output result, QPijRepresenting the output result of the twin simulation path model, wherein the two results are represented in a two-dimensional matrix mode;
s42: generating ganLoss as a function of the countermeasure loss, wherein the formula is as follows:
wherein, PmodelWhen the counterstudy model is used for counterstudy, the counterstudy model considers a probability matrix that the result output by the twin target object model is completely consistent with the result output by the twin simulation path model, the matrix is a two-dimensional matrix and accommodates different matrixesP of roads with different time periodsmodelA value; the EPCG is a dimension stability factor when the gradient is reduced, and the EPCG adopts a pseudorandom function in the simulation of [10 ]-4,10-11]The interval takes one value in a pseudo-random manner.
S43: and (3) adopting a discrimination formula to discriminate the random function, wherein the smaller the loss function is, the more accurate the simulation result is, wherein the discrimination formula is as follows:
wherein,
is the desired probability, i.e. the probability matrix whose actual situation is consistent with the result output by the twin target object model.
The whole process of establishing the models can be summarized into the following stages:
the first stage is as follows: and obtaining a twin target object characteristic model by a random forest method.
And a second stage: establishing a twin simulation path model based on a Bayesian network according to the twin target object characteristic model, and simulating to output a traffic situation evaluation result; and establishing a twin target object simulation model based on the modified Agent model according to the twin target object characteristic model to simulate and output a traffic situation evaluation result.
And a third stage: based on the actual traffic situation, the twin simulation path model and the twin target object simulation model perform antagonistic learning, and the importance of the twin target object characteristic model is modified, so that the simulation is closer to the actual traffic situation, and the core comment loss function is established as follows:
a fourth stage: and selecting the model with the minimum loss function, carrying out traffic state simulation, and outputting the result.
And the data output submodule is used for sending the road traffic situation evaluation result to the communication module and the visualization module.
The part of a core model submodule of the calculation analysis module is a small-sized calculation unit (a microcomputer or other calculation units with similar functions), the part stores a road traffic situation assessment simulation model based on the digital twin, and after data is read and preprocessing is executed, the road traffic situation assessment simulation model based on the digital twin is called to simulate the traffic situation in real time, so that the method is very accurate and efficient. The data reading submodule reads a target object and transmits data of the environment where the target object is located to the data preprocessing submodule, the data preprocessing submodule preprocesses the format of the data and transmits the data to the core model submodule, the core model submodule corrects the data according to the preprocessed data, the twin simulation path model and the twin target object simulation model are established, self-optimization is carried out based on a counterstudy method, an optimal model is selected for simulation evaluation, an evaluation result is generated, the result is transmitted to the data output submodule, the data output submodule converts the data into a format which is suitable for the visualization module to carry out data visualization, and finally the data is transmitted to the data visualization module, so that the working closed loop of the calculation analysis module is completed.
The calculation analysis module has universality, and can call a road traffic situation assessment simulation model based on the digital twin by adopting the same strategy aiming at different target objects (including various types of vehicles), so that the simulation method is ensured to be suitable for completing various tasks.
The road traffic situation assessment simulation model based on the digital twin in the calculation analysis module is formed by preferably establishing an advanced agent framework based on a widely recognized intelligent modeling platform analog, adjusting detailed numerical values (such as the number of model layers, the number of units and the like) to obtain a target object simulation model with good efficiency and performing antagonistic learning with a twin simulation path model established based on a Tensorflow platform.
The software part of the calculation analysis module comprises: the data reading program reads the relevant data of the target object and the relevant data of the road section where the target object is located from an information extraction module; the data preprocessing program transmits data to the core model unit after necessary preprocessing such as interception, format conversion and the like; and the data output program transmits the simulation result of the core model unit to the visualization module and the communication module.
As an implementation mode, the signal transmission mode of the communication module can be a common WIFI or Bluetooth transmission mode, and the two modes are large in information transmission amount and high in speed.
As a rule, the traffic situation assessment simulation result sent by the calculation and analysis module is subjected to data conversion to become information which can be displayed on an electronic screen, the traffic situation assessment simulation result is displayed in real time in a color, digital, curve or graphic mode and the like, and the results are displayed in a video data or picture data mode.
Through the data extraction module, the calculation analysis module, the communication module, the visualization module and the additional module, a hardware carrier and capability support are provided for the traffic situation assessment simulation, and real-time data acquisition and storage simulation are carried out.
In a specific evaluation model, as shown in fig. 5, the target object is a traveling vehicle in a certain road section, and the environment of the target object is as shown in the figure. In the embodiment of the invention, the information extraction module extracts a series of data such as the length, the width, the average speed and the like of the target object and a series of data such as the length, the width, the weather, the peak time ratio and the like of the road section where the target object is located, and generates a data packet as a basis for subsequent modeling.
And when the information extraction module extracts data, the calculation analysis module is activated, the data of the sensing information extraction module is continuously read through the analog platform program, necessary preprocessing is carried out, and the necessary preprocessing is provided for a core prediction model in the calculation analysis module. As shown in fig. 5, the core prediction model is obtained by firstly screening important feature attributes through a random forest model, forming an important feature data set, establishing a feasible road section as W01 and a driving vehicle as E01 in a twin simulation path model and twin target object simulation model diagram, performing antagonistic learning through the twin simulation path model and the twin target object simulation model, and continuously optimizing after correcting the importance of the features of the twin target object. After multiple times of simulation counterwork, an optimal model is selected for simulation evaluation, and a predicted dynamic traffic situation map is obtained. The number of vehicles, the locations where the vehicles are present, and the flow direction of the vehicles reflect the traffic situation synchronously.
When the calculation analysis module develops prediction and gives a result in real time, the built-in java program transmits the result to the communication module and the visualization module at the same time, and the communication module directly transmits the evacuation simulation result to the designated server according to a preset instruction.
The information extraction module is used for obtaining the information of the target object, the calculation analysis module is used for providing resource (calculation force) support for analysis, the communication module is used for providing result transmission capability, and the visualization module is used for providing a result display platform. The device can accurately predict the traffic situation assessment condition of the target road section in real time, continuously updates the result along with the evolution of the situation, and has important significance on the work of traffic situation prediction, control and the like.
The data information of the target object and the space where the target object is located extracted by the extraction module is read and preprocessed by the calculation analysis module, the preprocessed data information of the target object and the space where the target object is located is subjected to analog analysis by the digital twin model to generate a road condition evaluation result, the road condition evaluation result is sent to the pre-receiving end by the communication module, and the road condition evaluation result is visualized by the visualization module. The method is accurate and efficient, can continuously update results along with the evolution of the situation, is obviously superior to the traditional traffic situation assessment method, and has great significance for the prediction, control and scheduling of traffic events and the planning of a traffic network.
Claims (7)
1. A road traffic situation simulation method based on digital twins is characterized by comprising the following steps:
s001: the calculation analysis module acquires the information of the target object and the information of the space where the target object is located through the information extraction module;
s002: reading and preprocessing twin target object characteristic information, modeling by a calculation analysis module through a digital twin-based modeling method, completing model self-optimization, selecting an optimal model for simulation, and generating a road situation evaluation result;
s003: sending the road situation evaluation result to a preset receiving end through a communication module;
s004: and visually converting the road situation evaluation result, and generating visual video data on a visual module in a high frame rate video mode.
2. The digital twin-based road traffic situation simulation method according to claim 1, characterized in that: in S002, the model created by the digital twinning-based modeling method includes a twinning target object feature model, a twinning target object model, and a twinning simulation path model.
3. The digital twin-based road traffic situation simulation method according to claim 2, characterized in that: the twin target object feature model is built by adopting a random forest modeling method and comprises the following steps,
s101, a calculation analysis module acquires road characteristic data after noise reduction and fusion of traffic road conditions through an information extraction module;
s102: the calculation analysis module reads and preprocesses the noise-reduced and fused road characteristic data of the traffic road condition, and carries out modeling analysis on the preprocessed noise-reduced and fused road characteristic data through a random forest model to generate an importance sequencing result;
s103, the calculation analysis module provides road characteristic data with the importance lower than 5%, and outputs a twin target object characteristic result;
and S104, storing the obtained twin target object characteristic result into a data set in a matrix form to form a twin target object characteristic model.
4. The digital twin-based road traffic situation simulation method according to claim 3, characterized in that: in S102, the modeling analysis of the random forest model includes the steps of,
s11: averagely dividing the total tree number M in the forest into 4 parts, distributing the 4 parts to each process for sharing, creating M/4 decision trees by each process, and returning the created M/4 decision trees to the main process in a list form;
s12: after obtaining the decision tree lists of 4 sub-processes, the main process integrates the 4 sub-lists into a decision tree list L with the length of M;
s13: generating a training set for each decision tree, and randomly extracting n samples from the original training sample set by using a bootstrapping aggregation method in bagging as a group of training sets;
s14: constructing a single decision tree, randomly extracting m features from the features of each sample, dividing the m features into a sample set B in a feature-based mode to calculate a Keyny index, and setting the sample set as a sample set D1 equal to a specified feature and a sample set B2 not equal to the given feature, then:
wherein A is the current characteristic, then finding out the division A with the minimum Gini index from all Gini (B, Ai), using the characteristic A as the splitting characteristic, and repeating the step to obtain a decision tree;
s15: each process repeatedly constructs M/3 decision trees according to the steps S13 and S14, the decision trees are integrated to form a random forest, and the last leaf node of each decision tree is the feature with the largest influence on network security to form a feature set A1 and A2 … … Am;
s16: judging the effectiveness of the model according to the accuracy and selecting the optimal parameters:
wherein TP is true positive, which indicates that the actual category is positive, and the algorithm output category is also positive; FP is false positive, which means that the actual category is negative, and the sample output category is positive; TN is true negative, which means that the actual category is negative, and the algorithm output category is also negative; FN is false negative, which indicates that the actual category is positive, and the algorithm output category is negative; the hyper-parameter θ is set to: max _ features is 0.2 and decision tree number tree is 18.
5. The digital twin-based road traffic situation simulation method according to claim 2, characterized in that: the twin target object model is established by adopting a method of modifying original Agent, comprising the following steps,
s31: an initialization stage: the vehicle physical entity is initialized to Agent, and the attribute of the Agent is defined as:
wherein id represents the unique identification number of the Agent;
representing a velocity vector of the Agent; pos (t) represents the location of Agent; h represents the length of the Agent; r represents the width of the Agent; p represents the average speed of the Agent; t represents a time step;
s32: and (3) an interaction stage: each Agent carries out information interaction with other agents and the environment where the agents are located, and the current position is recorded in a coordinate mode;
s33: an action stage: each Agent moves according to the driving of the social force model, and continuously updates the position information until the action stops;
s34: and recording and returning the speed change condition of the action on the road section, and judging and returning the traffic delay condition of the road according to the occupation ratio of the agent action low-speed time section in the running time section.
6. The digital twin-based road traffic situation simulation method according to claim 4, characterized in that: the twin simulation path model is built and trained by adopting a Bayesian network model, and comprises the following steps,
s21: features obtained by random forest model learning simulationSign sample set { A1，A2……AmCombining the occurrence probability with the corresponding occurrence time period and a sample label to form a feature set sample T with an occurrence probability value, recording the occurrence probability as prior probability, updating the feature sample, correcting all the prior probabilities by using a Bayes formula, and solving the posterior probability;
s22: taking parameters related to the road section traffic situation as nodes, wherein the nodes comprise observable nodes and hidden nodes; the observable nodes comprise the average travel time of the road sections and the relative density of the road sections, which can cause traffic, and the hidden nodes comprise the traffic situation of the road sections and the historical probability of traffic accidents of the road sections;
s23: calculating the probability of traffic delay of the corresponding road section by adopting a Bayesian formula, wherein the Bayesian formula is as follows:
where P (A | B) is the likelihood of A occurring in the case of B occurring, A1, … …, Ai are complete event groups, P (A) is the prior probability or marginal probability of A, P (A | B) is the conditional probability of A after B is known to occur, P (B | A) is the conditional probability of B after A is known to occur, and P (B) is the prior probability or marginal probability of B.
S24: and forming a Bayesian network for predicting the traffic delay condition of the road, and outputting a result.
7. The digital twin-based road traffic situation simulation method according to claim 6, characterized in that: taking the output result of the Bayesian network of S24 as the result of the simulation of the object for the twin target object model to carry out the antagonistic learning comprehensive review, establishing an antagonistic learning model,
the result output by the twin target object model is reviewed by adopting the result output by the twin simulation path model, and the result is evaluated according to a comprehensive evaluation method, which comprises the following steps,
s41: a loss function gloss is generated, and the formula is:
loss＝ganWeight×ganLoss+modelWeight×modelLoss；
wherein the ganWeight is the weight of the loss-resisting function ganLoss, and the value is 0-30; the modelWeight is the weight of a loss function modelLoss of a difference value between a twin target object model output result and a twin simulation path model output result, and the value of the modelWeight is 0-50;
modelLoss represents the difference between the output result of the twin target object model and the output result of the twin simulation path model, and the calculation formula is as follows:
wherein, RPijRepresenting the twin target object model output result, QPijRepresenting the output result of the twin simulation path model, wherein the two results are represented in a two-dimensional matrix mode;
s42: generating ganLoss as a function of the countermeasure loss, wherein the formula is as follows:
when the Pmodel is confrontation learning, the confrontation learning model considers a probability matrix that a result output by the twin target object model is completely consistent with a result output by the twin simulation path model, the matrix is a two-dimensional matrix and contains Pmodel values of different roads in different time periods; the EPCG is a dimension stability factor when the gradient is reduced, and the EPCG adopts a pseudorandom function in the simulation of [10 ]-4,10-11]Taking a value in the interval in a pseudo-random mode;
s43: and (3) adopting a discrimination formula to discriminate the random function, wherein the smaller the loss function is, the more accurate the simulation result is, wherein the discrimination formula is as follows:
wherein,
is the desired probability, i.e. the probability matrix whose actual situation is consistent with the result output by the twin target object model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210099127.8A CN114519302B (en) | 2022-01-25 | 2022-01-25 | Highway traffic situation simulation method based on digital twinning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210099127.8A CN114519302B (en) | 2022-01-25 | 2022-01-25 | Highway traffic situation simulation method based on digital twinning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114519302A true CN114519302A (en) | 2022-05-20 |
| CN114519302B CN114519302B (en) | 2024-06-28 |
Family
ID=81596143
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210099127.8A Active CN114519302B (en) | 2022-01-25 | 2022-01-25 | Highway traffic situation simulation method based on digital twinning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114519302B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114943304A (en) * | 2022-06-17 | 2022-08-26 | 广东工业大学 | Bayesian-based 3C electronic product assembly error transfer method |
| CN115481531A (en) * | 2022-09-06 | 2022-12-16 | 苏州大学 | Road network traffic flow real-time twinning method and system based on SUMO |
| CN115718466A (en) * | 2022-11-21 | 2023-02-28 | 广东工业大学 | Digital twin workshop fault prediction method based on random forest and analytic hierarchy process |
| CN115952692A (en) * | 2023-03-10 | 2023-04-11 | 腾讯科技（深圳）有限公司 | Road traffic simulation method and device, storage medium and electronic equipment |
| CN117057656A (en) * | 2023-08-17 | 2023-11-14 | 广东飞翔云计算有限公司 | Digital twinning-based smart city management method and system |
| CN118196662A (en) * | 2024-03-26 | 2024-06-14 | 鄂尔多斯市腾远煤炭有限责任公司 | Coal mine digital twin model building method based on Bayesian network algorithm |
| CN118410657A (en) * | 2024-07-03 | 2024-07-30 | 北京国星创图科技有限公司 | Satellite measurement and control interactive simulation and emission management integrated method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111191683A (en) * | 2019-12-13 | 2020-05-22 | 南京邮电大学 | Network security situation assessment method based on random forest and Bayesian network |
| AU2020100709A4 (en) * | 2020-05-05 | 2020-06-11 | Bao, Yuhang Mr | A method of prediction model based on random forest algorithm |
| CN112837344A (en) * | 2019-12-18 | 2021-05-25 | 沈阳理工大学 | Target tracking method for generating twin network based on conditional confrontation |
| CN113704956A (en) * | 2021-06-15 | 2021-11-26 | 深圳市综合交通设计研究院有限公司 | Urban road online microscopic simulation method and system based on digital twin technology |
-
2022
- 2022-01-25 CN CN202210099127.8A patent/CN114519302B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111191683A (en) * | 2019-12-13 | 2020-05-22 | 南京邮电大学 | Network security situation assessment method based on random forest and Bayesian network |
| CN112837344A (en) * | 2019-12-18 | 2021-05-25 | 沈阳理工大学 | Target tracking method for generating twin network based on conditional confrontation |
| AU2020100709A4 (en) * | 2020-05-05 | 2020-06-11 | Bao, Yuhang Mr | A method of prediction model based on random forest algorithm |
| CN113704956A (en) * | 2021-06-15 | 2021-11-26 | 深圳市综合交通设计研究院有限公司 | Urban road online microscopic simulation method and system based on digital twin technology |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114943304A (en) * | 2022-06-17 | 2022-08-26 | 广东工业大学 | Bayesian-based 3C electronic product assembly error transfer method |
| CN114943304B (en) * | 2022-06-17 | 2023-04-07 | 广东工业大学 | Bayesian-based 3C electronic product assembly error transfer method |
| CN115481531A (en) * | 2022-09-06 | 2022-12-16 | 苏州大学 | Road network traffic flow real-time twinning method and system based on SUMO |
| CN115481531B (en) * | 2022-09-06 | 2023-07-11 | 苏州大学 | SUMO-based road network traffic flow real-time twin method and system |
| CN115718466A (en) * | 2022-11-21 | 2023-02-28 | 广东工业大学 | Digital twin workshop fault prediction method based on random forest and analytic hierarchy process |
| CN115718466B (en) * | 2022-11-21 | 2024-05-14 | 广东工业大学 | Digital twin workshop fault prediction method based on random forest and analytic hierarchy process |
| CN115952692A (en) * | 2023-03-10 | 2023-04-11 | 腾讯科技（深圳）有限公司 | Road traffic simulation method and device, storage medium and electronic equipment |
| CN117057656A (en) * | 2023-08-17 | 2023-11-14 | 广东飞翔云计算有限公司 | Digital twinning-based smart city management method and system |
| CN117057656B (en) * | 2023-08-17 | 2024-05-31 | 广东飞翔云计算有限公司 | Digital twinning-based smart city management method and system |
| CN118196662A (en) * | 2024-03-26 | 2024-06-14 | 鄂尔多斯市腾远煤炭有限责任公司 | Coal mine digital twin model building method based on Bayesian network algorithm |
| CN118410657A (en) * | 2024-07-03 | 2024-07-30 | 北京国星创图科技有限公司 | Satellite measurement and control interactive simulation and emission management integrated method and system |
| CN118410657B (en) * | 2024-07-03 | 2024-09-10 | 北京国星创图科技有限公司 | Satellite measurement and control interactive simulation and emission management integrated method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114519302B (en) | 2024-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114519302B (en) | Highway traffic situation simulation method based on digital twinning | |
| CN111061277B (en) | Unmanned vehicle global path planning method and device | |
| CN109508360B (en) | Geographical multivariate stream data space-time autocorrelation analysis method based on cellular automaton | |
| CN110164128B (en) | City-level intelligent traffic simulation system | |
| WO2023207437A1 (en) | Scene flow digital twin method and system based on dynamic trajectory flow | |
| CN111612206B (en) | Neighborhood people stream prediction method and system based on space-time diagram convolution neural network | |
| CN110443969B (en) | Fire detection method and device, electronic equipment and storage medium | |
| CN106815563B (en) | Human body apparent structure-based crowd quantity prediction method | |
| CN114332473A (en) | Object detection method, object detection device, computer equipment, storage medium and program product | |
| CN113033899A (en) | Unmanned adjacent vehicle track prediction method | |
| CN118297775B (en) | Urban planning management and control system based on digital twin technology | |
| Lin et al. | Optimal CNN-based semantic segmentation model of cutting slope images | |
| Chen et al. | Advanced driver assistance strategies for a single-vehicle overtaking a platoon on the two-lane two-way road | |
| CN117911829B (en) | Point cloud image fusion method and system for vehicle navigation | |
| Rong et al. | Big data intelligent tourism management platform design based on abnormal behavior identification | |
| CN116823572B (en) | Population flow data acquisition method and device and computer readable storage medium | |
| CN116597419B (en) | Vehicle height limiting scene identification method based on parameterized mutual neighbors | |
| KR102014288B1 (en) | Development pressure prediction method based on artificial intelligence using drone | |
| Byju et al. | Pedestrian detection and tracking in challenging conditions | |
| CN115981302A (en) | Vehicle following lane change behavior decision-making method and device and electronic equipment | |
| CN115565388A (en) | Traffic light control method based on multi-channel vehicle detection and three-dimensional feature labeling | |
| Jenifer et al. | Improved Mayfly Optimization and LightGBM Classifier for Smart City Traffic Prediction | |
| Mohammed et al. | Toward agent-based microsimulation of cyclist following behavior: Estimation of reward function parameters using inverse reinforcement learning | |
| CN116311950B (en) | Path selection method and V2X test system based on virtual-real fusion technology | |
| CN117132958B (en) | Road element identification method and related device |
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 |