CN116805449A - Traffic signal fuzzy control method and system considering individual priori data - Google Patents
Traffic signal fuzzy control method and system considering individual priori data Download PDFInfo
- Publication number
- CN116805449A CN116805449A CN202310815824.3A CN202310815824A CN116805449A CN 116805449 A CN116805449 A CN 116805449A CN 202310815824 A CN202310815824 A CN 202310815824A CN 116805449 A CN116805449 A CN 116805449A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- fuzzy
- green light
- phase
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 72
- 238000005457 optimization Methods 0.000 claims abstract description 13
- 238000012706 support-vector machine Methods 0.000 claims abstract description 13
- 238000011897 real-time detection Methods 0.000 claims abstract description 6
- 238000004140 cleaning Methods 0.000 claims abstract description 4
- 230000007704 transition Effects 0.000 claims description 17
- 238000011144 upstream manufacturing Methods 0.000 claims description 16
- 230000008569 process Effects 0.000 claims description 9
- 238000011160 research Methods 0.000 claims description 8
- 238000013461 design Methods 0.000 claims description 7
- 238000001514 detection method Methods 0.000 claims description 7
- 238000005070 sampling Methods 0.000 claims description 7
- 230000005484 gravity Effects 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 4
- 238000013398 bayesian method Methods 0.000 claims description 3
- 230000007423 decrease Effects 0.000 claims description 3
- 230000009977 dual effect Effects 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 3
- 238000012806 monitoring device Methods 0.000 claims 1
- 238000004364 calculation method Methods 0.000 abstract description 9
- 238000004088 simulation Methods 0.000 description 11
- 230000000694 effects Effects 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000011217 control strategy Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 206010021033 Hypomenorrhoea Diseases 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- 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
- Traffic Control Systems (AREA)
Abstract
The invention relates to a traffic signal fuzzy control method considering individual priori data, which comprises the following steps: step 1, cleaning and analyzing license plate identification data; step 2, performing personalized travel time prediction of a support vector machine driver based on Bayesian optimization; step 3, performing Bayesian driver personalized path prediction based on Markov chains; step 4, fusing and utilizing the historical trip information and the real-time detection information of the vehicle; and 5, obtaining the phase arrival rate and the turning rate through vehicle travel time and path prediction, and constructing a traffic signal fuzzy control method. The traffic flow arrival rate calculation thought provided by the invention can effectively estimate the traffic flow arrival rate under four-phase control, and the calculation result can be used as a data source for traffic signal control; the average delay of the vehicle can be obviously reduced, the high-precision prediction of the travel time of the vehicle is realized, and the service quality of the intersection is obviously improved.
Description
Technical Field
The invention belongs to the technical field of traffic engineering, and particularly relates to a traffic signal fuzzy control method and system considering individual priori data.
Background
The fuzzy logic signal control is used for timing the traffic signal control through the fuzzy logic control, is one of the most classical and most effective methods in the traffic signal control field, can effectively optimize the signal timing result, and has remarkable effects of relieving intersection congestion and improving intersection operation efficiency. The existing signal control method based on fuzzy logic control is mainly used for optimizing signal timing only based on real-time detection data, is lack of utilization of historical data and predicted data, is long in calculation time, causes serious hysteresis of a control strategy, causes traffic jam, and cannot realize real-time control.
With the development of intelligent traffic detection technology, individual vehicle travel time-space data can be obtained more easily, and the mining of individual travel data provides a new solution idea and method for improving traffic signal control effect. Therefore, there is an urgent need for a traffic signal fuzzy control method and system that considers individual prior data, and by comprehensively utilizing vehicle history data and real-time data, the time for vehicles to reach an intersection is predicted, and the traffic efficiency of the intersection is effectively improved.
Disclosure of Invention
The invention aims to solve the problem that the current intersection adaptive signal control mode lacks comprehensive utilization of vehicle historical data and real-time data, so that a control strategy is seriously lagged to cause traffic jam, and further provides a traffic signal fuzzy control method considering individual priori data.
The technical scheme adopted by the invention for solving the technical problems is as follows: a traffic signal fuzzy control method considering individual prior data comprises the following steps:
step 1, cleaning and analyzing license plate identification data;
step 2, performing personalized travel time prediction of a support vector machine driver based on Bayesian optimization;
step 3, performing Bayesian driver personalized path prediction based on Markov chains;
step 4, fusing and utilizing the historical trip information and the real-time detection information of the vehicle;
and 5, obtaining the phase arrival rate and the turning rate through vehicle travel time and path prediction, and constructing a traffic signal fuzzy control method.
Further, in the step 1, the historical trip data of the individual vehicles are obtained through a traffic detection technology, a historical trip database is constructed, and the historical trip data are cleaned and analyzed.
In step 2, a support vector machine model is adopted to predict the travel time of the driver, and the weather, the time period, the average travel time of the road section and the average travel time characteristics of the driver are utilized to predict the travel time of the driver in the future time period; x is x i To influence the factors of travel time prediction, x j As a predicted value of the travel time, a radial basis function and a decision function of the support vector machine travel time prediction model are as follows:
k(x i ,x j )＝exp(-γ||x i -x j || 2 ) (1)
wherein alpha is i For the solution of the dual problem converted by utilizing the Lagrangian multiplier method, gamma is the bandwidth of a kernel function, and b is a bias term;
in the process of predicting the travel time by using the model, searching an optimal parameter combination of the radial basis function bandwidth parameter gamma, the penalty coefficient C and the insensitive coefficient epsilon by adopting a Bayesian optimization method, namely calculating posterior probability by using prior probability and observation data to perform parameter optimization;
the proxy model is as follows:
the mean function is μ (x), the covariance function is k (x, x'), and the Gaussian process is expressed as
f(x)～GP(μ(x),k(x,x′)) (3)
The sampling strategy is used for determining the value of the next set of super parameters, and the used sampling strategy is expected to be improved, and the formula is as follows:
wherein μ (x) is the posterior distribution mean, σ (x) is the covariance, f (x) t+1 ) Is an optimal objective function.
Further, in the step 3, a bayesian method is used for predicting the vehicle path, real-time license plate recognition is realized according to the bayonet monitoring equipment, the partial running path of the vehicle is determined, and the vehicle front path T is obtained q Predicting the position of the vehicle passing through the next bayonet, converting the problem into calculating a known preamble path T based on a Bayesian model q Under the condition that the vehicle passes through the next intersection l j Probability of (2):
wherein: p (l) j ) For the prior probability, the position of the next bayonet of the vehicle is l j The probability of passing through the bayonet l j A ratio of the number of paths to the amount of historical path data;
in the process of predicting the travel path of the individual vehicle, the travel path is regarded as a Markov chain, the next position of the driver is considered to be related to the current position of the driver, and the primary travel path T= { T i ,L,t j },t k E R decomposition into
T＝{t i ,t i+1 }{t i+1 ,t i+2 }L{t j-1 ,t j },t k ∈R
Individual vehicle passing path t= { T i ,t i+1 }{t i+1 ,t i+2 }L{t j-1 ,t j },t k After e R, the probability of going from Bayonet li to Bayonet lj is as follows:
P(T)＝p i,i+1 p i+1,i+2 L p j-1,j (6)
wherein: p is p ij For two adjacent bayonets l i ，l j The state transition probability of (1) corresponds to the state transition probability of the interface i Driving direction bayonet l j The probability of (1) is that it contains l i -l j The number of historical paths and the inclusion of l i A ratio of the number of historical paths;
constructing a state transition probability matrix M between all adjacent bayonets:
the total path probability between two bayonets is as follows:
wherein: s is the shortest distance between bayonets i, j; k is the detour coefficient
When the vehicle passes through the bayonet j, the front path is T q Probability of (2):
wherein P (T) q ) Is a path probability; p is p c→j The transition probability of the vehicle reaching the next bayonet j from the current position is given; p is p s→j The transition probability of the vehicle from the starting position s to the next bayonet j;
obtaining the vehicle on the known path T according to the formula (1) q Under the condition of (1), the probability of passing through different bayonets at the next moment, and the bayonets with the highest probability are selected as path prediction results.
Further, in the step 4, the traffic detection device extracts the historical travel information of the vehicle, and constructs a personalized travel characteristic database of the vehicle, including travel mode and route preference information of the vehicle; comparing the real-time detected vehicle information with a historical database to identify and verify the identity and behavior of the vehicle, and predicting the travel time of the vehicle with high precision by using a travel time prediction model in combination with the current weather, time period, road section average travel time and driver average travel time to predict the travel time of the vehicle from the upstream to the target intersection; predicting the next occurrence position of the vehicle through a vehicle path prediction model according to the travel path track, analyzing the steering selection behavior of the vehicle, obtaining the phase required by the vehicle to pass through an intersection, and predicting the next occurrence position information and occurrence time of the vehicle; the arrival prediction information of the vehicles is corresponding to the signal period and the phase of the intersection, the arrival vehicles of each phase in each period are accumulated and summed, and the arrival vehicles are divided by the period duration to obtain the predicted arrival rate of each phase of the intersection in each period; and after the travel is finished, updating a vehicle personalized travel characteristic database by using the data detected in real time, and analyzing the behavior mode of the individual vehicle.
Further, in the step 5, based on the four-phase setting, the traffic demand of each flow direction is predicted by combining the data of the detector at the upstream of the intersection, and the vehicle travel time is reduced as a target, so as to construct the traffic signal control method based on the two-stage fuzzy controller.
Initializing a signal control system, predicting the arrival rate according to the green light phase, and outputting preliminary green light delay by using a first-stage fuzzy controller; the first-stage fuzzy controller is a current green light phase traffic intensity judging module, and the module preliminarily outputs green light extension time according to the arrival condition of vehicles in green light phases, and the design steps are as follows:
(1) Input and output of fuzzy controller
Input variables: predicted arrival rate lambda of current green light phase gi
Output variable: green light delay DT 1
(2) Variable domain of discussion and scale factor
The upstream detector is arranged at the starting point of a section of connection between the upstream intersection and the target intersection, and the downstream detector is arranged at the stop line of the entrance road of the target intersection; according to the actual traffic flow condition of the research area and related research, the current predicted arrival rate lambda of the green light phase gi Is set to [0,0.3 ]]The scale factor takes 1, and the fuzzy universe is [0,0.3]The method comprises the steps of carrying out a first treatment on the surface of the Green light delay DT 1 Basic discourse domain settingIs [0,30]The scale factor takes 1, and the fuzzy universe is [0,30]
(3) Fuzzy subset of input-output variables
Predicted arrival rate lambda of current green light phase gi Dividing into 7 fuzzy subsets, respectively { very low, medium, high, very high }, denoted { VL, L, RL, M, RH, H, VH }; all input and output variables of the fuzzy controller are divided into 7 fuzzy subsets
(4) Membership function
Selecting a triangle membership function
(5) Control rules
The first-stage fuzzy controller is a single-input single-output fuzzy controller, and the traffic intensity of the current green light phase, namely the predicted arrival rate lambda of the vehicle gi The larger the extension time corresponding to the green light phase is, the longer the extension time is; conversely, the predicted arrival rate lambda of the vehicle at the current green light phase gi The smaller the extension time corresponding to the green light phase is, the shorter the extension time is; 7 corresponding rules are formulated according to fuzzy subset division of the input and output variables;
(6) Deblurring
The gravity center method is used for resolving ambiguity, all functions of the primary fuzzy controller are achieved through programming, and input variables are set: predicted arrival rate lambda of current green light phase gi And obtaining an output variable after fuzzification, fuzzy reasoning and fuzzy operation: green light delay DT 1 Is a value of (a).
After the target phase is converted into a red light, the red light enters a second-stage fuzzy controller, the arrival rate is predicted according to the red light phase, and the second-stage fuzzy controller is used for outputting green light delay compensation; the second-stage fuzzy controller is a current red-light phase traffic intensity judging module, the module judges the traffic intensity of the red-light phase from the current phase to the initial green-light phase through fuzzy reasoning according to the green-light extension time output by the first-stage fuzzy controller according to the arrival condition of the current three red-light phase vehicles, and outputs green-light delay compensation time, and the design steps are as follows:
(1) Input and output of fuzzy controller
Input variables: green light delay DT 1 Maximum predicted arrival rate lambda of current red light phase ri
Output variable: green light delay compensation DT 2
(2) Variable domain of discussion and scale factor
Maximum predicted arrival rate lambda of current red light phase ri Is set to [0,0.3 ]]The scale factor takes 1, and the fuzzy universe is [0,0.3]The method comprises the steps of carrying out a first treatment on the surface of the Green light delay DT 1 The basic domain is set to [0,30]Due to the delay compensation time DT 2 Cannot exceed the green light delay DT 1 So green light delay DT 1 The scale factor of (2) is taken as 2, and the fuzzy discourse domain is [0,60 ]]The method comprises the steps of carrying out a first treatment on the surface of the Green light delay compensation DT 2 The basic domain is set to [0,30]The scale factor takes 1, and the fuzzy universe is [0,30]
(3) Fuzzy subset of input-output variables
Maximum predicted arrival rate lambda of current red light phase ri Green light delay DT 1 Green light delay compensation DT 2 Are all divided into 7 fuzzy subsets
(4) Membership function
Consistent with the first-stage fuzzy controller, a triangle membership function is selected
(5) Control rules
The second-stage fuzzy controller is a multi-input single-output fuzzy controller, and predicts the arrival rate lambda ri The larger the green light delay DT 1 The longer the delay compensation time DT 2 The longer; conversely, the predicted arrival rate lambda ri Smaller green light delay DT 1 The shorter the delay compensation time DT 2 The shorter; according to fuzzy subset division of input and output variables, 49 corresponding rules are formulated
(6) Deblurring
Adopting a gravity center method to deblur, and obtaining a deblurred output curved surface through programming; for lambda ri And DT (DT) 1 Can find the unique correspondence in the defuzzified curved surface and along with lambda ri And DT (DT) 1 Is reduced, green light delay compensated DT 2 And also decreases with lambda ri And DT (DT) 1 Is increased, green light delay compensation DT 2 And also increases with it；
And outputting a green lamp delay according to the actual situation of the intersection, and executing the next phase.
The invention also relates to a system of the traffic signal fuzzy control method taking the individual prior data into consideration, which comprises a computer module of the traffic signal fuzzy control method taking the individual prior data into consideration. The computer module comprises a first-stage fuzzy controller, a second-stage fuzzy controller, a decision module and a phase switching module.
Advantageous effects
The invention mainly provides a traffic signal fuzzy control method considering individual priori data, which is characterized in that the traffic signal control method based on a secondary fuzzy controller is constructed by predicting the travel time of a vehicle from the upstream to a target intersection, periodically matching the vehicle with signals and predicting the travel path of the vehicle to obtain the phase required by the vehicle passing through the intersection, and further providing a predicted arrival rate calculation method of an entrance road of the intersection.
The traffic flow arrival rate calculation thought provided by the invention can effectively estimate the traffic flow arrival rate under four-phase control, and the calculation result can be used as a data source for traffic signal control; in addition, the invention can obviously reduce the average delay of the vehicle, compared with the traditional fuzzy control, the average delay of the vehicle is reduced by 18.75 percent in the peak leveling period, the average delay of the vehicle is reduced by 16.11 percent in the peak period, the service quality of an intersection is obviously improved, and important reference is provided for traffic signal control in the future whole-body vehicle sample environment.
Drawings
FIG. 1 is a schematic diagram of a control method according to the present invention;
FIG. 2 is a flowchart of a vehicle journey time prediction model algorithm based on a Bayesian parameter optimization support vector machine in the invention;
FIG. 3 is a flowchart of a vehicle path prediction algorithm based on a Bayesian method in the present invention;
FIG. 4 is a flow chart of a single intersection traffic signal fuzzy control design in the present invention;
FIG. 5 is a schematic view of an intersection used in the simulation process of the present invention;
FIG. 6a is a timing scheme of timing signal control during a flat-peak period simulation of the present invention;
FIG. 6b is a timing scheme for timing signal control during rush hour simulation according to the present invention;
FIG. 7a is a schematic diagram of the placement of the detector in scenario 2 during a simulation of the present invention;
FIG. 7b is a schematic diagram of the placement of the detector in scheme 3 during simulation of the present invention.
Detailed Description
Embodiments of the present invention will be specifically described below with reference to fig. 1 to 4.
The invention relates to a traffic signal fuzzy control method considering individual priori data, which mainly comprises the following steps:
step 1: cleaning and analyzing license plate identification data
Firstly, extracting data fields from license plate identification data, wherein the data fields comprise an intersection name, a road section name, intersection longitude and latitude, passing time, driving direction and the like; then, carrying out data protocol, analyzing abnormal conditions of data and carrying out data preprocessing; then analyzing the preprocessed data, wherein the preprocessed data comprises the number of days to be detected, the number of bayonets passing through each track and the time distribution of vehicles passing through two bayonets; and finally, developing and researching the time and space characteristics of vehicle travel in the researched area, dividing the early and late peak time of vehicle travel, and determining the area with larger vehicle travel flow and the road and intersection with large vehicle flow.
Step 2: support vector machine driver personalized travel time prediction based on Bayesian optimization
As shown in fig. 2, based on the historical trip database obtained in step 1, vehicle road section travel time data is extracted, and a bayesian optimization support vector machine is utilized. Support vector machine regression (SVR) is the mapping of data into gaussian space, finding an optimal hyperplane to keep samples of different classes as separated as possible in that space. And carrying out high-precision prediction on the travel time of the vehicle by combining influence factors such as weather, time period, average travel time of a road section, average travel time of a driver and the like, and generating a travel time prediction model of the vehicle from upstream to a target intersection.
Wherein, support vector machine (BO-SVM) method based on Bayesian optimization:
let x be i To influence the factors of travel time prediction, x j As a predicted value of the travel time, a radial basis function and a decision function of the support vector machine travel time prediction model are as follows:
k(x i ,x j )＝exp(-γ||x i -x j || 2 ) (10)
wherein alpha is i To solve the dual problem transformed by the Lagrangian multiplier method, γ is the bandwidth of the kernel function and b is the bias term.
In the process of predicting the travel time by using the model, an optimal parameter combination of the radial basis function bandwidth parameter gamma, the penalty coefficient C and the insensitive coefficient epsilon is sought by adopting a Bayesian optimization method, namely, the posterior probability is calculated by using the prior probability and the observation data to carry out parameter optimization.
(1) Proxy model
Assuming the mean function is μ (x) and the covariance function is k (x, x'), the gaussian process can be expressed as:
f(x)～GP(μ(x),k(x,x′)) (12)
(2) Sampling strategy
The sampling strategy is used for determining the value of the next set of super parameters, and the used sampling strategy is expected to be improved, and the formula is as follows:
wherein μ (x) is the posterior distribution meanSigma (x) is covariance, f (x) t+1 ) Is an optimal objective function.
Step 3: bayesian driver personalized path prediction based on Markov chain
As shown in fig. 3, based on the historical trip database obtained in step 1, individual vehicle trip path information is extracted, a markov transition matrix is constructed through the state transition characteristics of a markov chain, the transition rule of the vehicle between different roads is captured, and a bayesian vehicle path prediction model is established.
Among them, the Markov chain-based Bayes (Markov-Bayes) method:
the real-time license plate recognition is realized according to the bayonet monitoring equipment, and the partial driving path of the vehicle is determined, so that the front path T of the vehicle can be obtained q Based on this, predicting the position of the next bayonet that the vehicle is most likely to pass through in space, based on a Bayesian model, the problem can be converted into a calculation of a known preamble path T q Under the condition that the vehicle passes the next possible intersection l j Probability of (2):
wherein: p (l) j ) For the prior probability, the position of the next bayonet of the vehicle is l j The probability of passing through the bayonet l j The ratio of the number of paths to the amount of historical path data.
In predicting the travel path of an individual vehicle, the travel path is considered as a Markov chain, and the next position of the driver is considered to be related to the current position of the driver. Therefore, the primary travel path t= { T can be set i ,L,t j },t k E R is decomposed into T= { T i ,t i+1 }{t i+1 ,t i+2 }L{t j-1 ,t j },t k ∈R。
Individual vehicle passing path t= { T i ,t i+1 }{t i+1 ,t i+2 }L{t j-1 ,t j },t k After E R, by bayonet l i Driving direction bayonet l j The probability of (2) is as follows:
P(T)＝p i,i+1 p i+1,i+2 L p j-1,j (15)
wherein: p is p ij For two adjacent bayonets l i ，l j State transition probabilities of (1), i.e. by bayonet l i Driving direction bayonet l j The probability of (1) is that it contains l i -l j The number of historical paths and the inclusion of l i Is a ratio of the number of historic paths.
Constructing a state transition probability matrix M between all adjacent bayonets:
the total path probability between two bayonets is as follows:
wherein: s is the shortest distance between bayonets i, j; k is the detour coefficient.
Therefore, when the vehicle passes through the bayonet j, the front path is T q Probability of (2):
wherein P (T) q ) Is a path probability; p is p c→j The transition probability of the vehicle reaching the next bayonet j from the current position is given; p is p s→j The probability of a transition of the vehicle from the starting position s to the next gate j.
Thus, the vehicle can be found to be in the known path T according to the formula (1) q Under the condition of (1), the probability of passing through different bayonets at the next moment, and selecting the bayonets with the highest probability as the path prediction result.
To analyze the result of selecting the steering direction of the vehicle intersection, the position relationship between the intersection and the road section is determined, the road section is named based on the name of the intersection, the order of the intersection is the direction, and the adjacent intersection and the road section are analyzedConstructing a road network steering position retrieval table according to the relative positions of the directions; after the path prediction is completed, the path t= { T i ,L,t j },t k E, R is split to form a steering path sequence { t } i ,t i+1 ,t i+2 }{t i+1 ,t i+2 ,t i+3 }L{t j-2 ,t j-1 ,t j },t k E R, where { t } i ,t i+1 The path l the vehicle takes i -l i+1 And is about to arrive at intersection l i+1 ，l i+2 And analyzing the steering behavior of the vehicle crossing according to the road network steering position retrieval table for predicting the obtained next position.
Step 4: fusion utilization of vehicle history trip information and real-time detection information
The traffic detection equipment detects vehicle traffic information in real time, matches historical trip data, combines the influence factors such as current weather, time period, average travel time of road sections, average travel time of drivers and the like, predicts the travel time of the vehicle with high precision by using a travel time prediction model, and predicts the travel time of the vehicle from upstream to a target intersection; meanwhile, predicting the next appearance position of the vehicle through a vehicle path prediction model according to the travel path track, analyzing the steering selection behavior of the vehicle, obtaining the phase required by the vehicle to pass through the intersection, obtaining the next position information and appearance time of the predicted vehicle, corresponding the arrival prediction information of the vehicle to the signal period and the phase of the intersection, accumulating and summing the arrival vehicles of each phase in each period, dividing the period duration, and obtaining the predicted arrival rate lambda of each phase of the intersection in each period gi 。
Step 5: phase arrival rate and turning rate are obtained through vehicle travel time and path prediction, and a traffic signal fuzzy control method is constructed
Step 5-1: initializing a signal control system, and setting minimum green light time g of each phase min Maximum green time g max Green light extension time Δg; setting the current green light phase i of the intersection as an initial phase, and giving the initial phase a minimum green light time g i ＝g min ；
Step 5-2: as shown in fig. 4, when the vehicle arrives at the control intersection, judging whether the green light of the target phase is finished, if the green light is not finished, counting the vehicle into the predicted vehicle arrival rate lambda of the current green light phase gi Entering a first-stage fuzzy controller within 2s before the green light ends, and inputting the vehicle arrival rate lambda predicted by the current green light phase gi Output green light delay DT 1 The method comprises the steps of carrying out a first treatment on the surface of the Otherwise, the vehicle is counted into the current red light phase predicted vehicle arrival rate lambda gi ，DT 1 And vary according to the arrival rate.
The first-stage fuzzy controller is a current green light phase traffic intensity judging module, and the module mainly considers the arrival condition of a vehicle in a current green light phase and preliminarily outputs green light extension time, and the design steps are as follows:
(1) Input and output of fuzzy controller
Input variables: predicted arrival rate lambda of current green light phase gi 。
Output variable: green light delay DT 1 。
(2) Variable domain of discussion and scale factor
In the invention, an upstream detector is arranged at the starting point of a road section where an upstream intersection and a target intersection are connected, and a downstream detector is arranged at the stop line of an entrance way of the target intersection. According to the actual traffic flow condition of the research area and related research, the current predicted arrival rate lambda of the green light phase gi Is set to [0,0.3 ]]The scale factor takes 1, and the fuzzy universe is [0,0.3]. Green light delay DT 1 The basic domain is set to [0,30]The scale factor takes 1, and the fuzzy universe is [0,30]。
(3) Fuzzy subset of input-output variables
Predicted arrival rate lambda of current green light phase gi Dividing into 7 fuzzy subsets, respectively { very low, medium, high, very high }, denoted { VL, L, RL, M, RH, H, VH }; all input and output variables of the fuzzy controller are divided into 7 fuzzy subsets.
(4) Membership function
The triangle membership function has the advantages of low computational complexity, flexible parameters, easy adjustment and the like, and the triangle membership function is selected.
(5) Control rules
The first-stage fuzzy controller is a single-input single-output fuzzy controller, and the traffic intensity of the current green light phase, namely the predicted arrival rate lambda of the vehicle gi The longer the extension time corresponding to the green light phase is; conversely, the predicted arrival rate lambda of the vehicle at the current green light phase gi The smaller the time, the shorter the extension time corresponding to the green light phase. So 7 corresponding rules are formulated according to the fuzzy subset division of the input and output variables.
(6) Deblurring
The gravity center method is used for resolving ambiguity, all functions of the primary fuzzy controller are achieved through programming, and input variables are set: predicted arrival rate lambda of current green light phase gi And obtaining an output variable after fuzzification, fuzzy reasoning and fuzzy operation: green light delay DT 1 Is a value of (a).
Step 5-3: after the target phase is converted into red light, the target phase enters a secondary fuzzy controller, and the maximum vehicle arrival rate lambda predicted by the current red light phase is input gi Output green light delay compensation DT 2 . Minimum green time g at target phase i Before ending, outputting final green light delay deltag=DT 1 -DT 2 。
The second-stage fuzzy controller is a current red-light phase traffic intensity judging module, the module mainly considers the arrival condition of vehicles with current three red-light phases and the green light extension time output by the first-stage fuzzy controller, judges the traffic intensity of the red-light phase from the current phase to the initial green light extension time through fuzzy reasoning, and finally outputs green light delay compensation time, and the design steps are as follows:
(1) Input and output of fuzzy controller
Input variables: green light delay DT 1 Maximum predicted arrival rate lambda of current red light phase ri
Output variable: green light delay compensation DT 2
(2) Variable domain of discussion and scale factor
Maximum predicted arrival rate lambda of current red light phase ri Is set to [0,0.3 ]]The scale factor takes 1, and the fuzzy universe is [0,0.3]. Green light delay DT 1 The basic domain is set to [0,30]Due to the delay compensation time DT 2 Cannot exceed the green light delay DT 1 So green light delay DT 1 The scale factor of (2) is taken as 2, and the fuzzy discourse domain is [0,60 ]]. Green light delay compensation DT 2 The basic domain is set to [0,30]The scale factor takes 1, and the fuzzy universe is [0,30]。
(3) Fuzzy subset of input-output variables
Maximum predicted arrival rate lambda of current red light phase ri Green light delay DT 1 Green light delay compensation DT 2 Are divided into 7 fuzzy subsets.
(4) Membership function
Consistent with the first-level fuzzy controller, a triangle membership function is selected.
(5) Control rules
The second-stage fuzzy controller is a multi-input single-output fuzzy controller, and predicts the arrival rate lambda ri The larger the green light delay DT 1 The longer the delay compensation time DT 2 The longer; conversely, the predicted arrival rate lambda ri Smaller green light delay DT 1 The shorter the delay compensation time DT 2 The shorter. And according to the fuzzy subset division of the input and output variables, formulating 49 corresponding rules.
(6) Deblurring
And the obfuscation is performed by adopting a gravity center method, and an output curved surface for the obfuscation can be obtained through programming. For lambda ri And DT (DT) 1 Can find the unique correspondence in the defuzzified curved surface and along with lambda ri And DT (DT) 1 Is reduced, green light delay compensated DT 2 And also decreases with lambda ri And DT (DT) 1 Is increased, green light delay compensation DT 2 And also increases.
Step 5-4: judging whether the four phases are all operated in the signal period, if not, selecting the maximum vehicle arrival rate lambda from the red light phases which are not operated ri Corresponding red light phaseSetting the bit as the next green light phase; if all the signals are operated, the signal period is ended, the phase of the other three phases with the maximum arrival rate of the red light phase is switched, and the maximum vehicle arrival rate lambda in the red light period is obtained ri The corresponding red light phase is set to the next green light phase.
Effect verification
The intersection shown in fig. 5 is selected as an actual research case to verify the effectiveness of the invention for improving the traffic efficiency of the intersection, and a flow input scheme of two periods of flat peak and peak is designed as shown in table 1.
TABLE 1 flow entry scheme
In order to verify the effectiveness of the method, SUMO simulation software is adopted to compare the traffic signal fuzzy control based on the vehicle travel time and path prediction designed by the method with the fuzzy control of timing signal control and real-time detection of the arrival rate. The saturation flow rate of the left-turn lane is set to 1550 (pch/s), the saturation flow rate of the straight-run lane is set to 1650 (pch/s), and the four-phase arrangement mode is adopted for the phase unification. The experiment ensures that various control modes in the same period have flow input with the same size, and three benefit evaluation indexes of average delay, queuing length and parking times of vehicles are selected to evaluate the signal control mode. The experimental protocol was as follows:
(1) Timing signal control scheme
In the scheme 1, a time-division timing signal control mode is adopted to carry out simulation experiments on the researched intersections, and control parameters controlled by timing signals are calculated according to traffic data of flat peaks and peaks by adopting a Webster optimal cycle duration formula. Setting the shortest green lamp time of the phase to be 10s, setting the yellow lamp time to be 3s, setting the total red time to be 2s, and setting the starting loss time to be 3s. The timing scheme is shown in fig. 6a, 6b, where gi represents the green light phase, i=1, 2,3,4 represents the four phases, y represents the yellow light phase, and r represents the red light phase.
(2) Fuzzy control scheme based on detection arrival rate
In the scheme 2, a fuzzy control mode of traffic signals with real-time arrival rate detected by coils is adopted, a detector is respectively arranged at the beginning of a guiding lane of each entrance lane, namely 50 meters upstream of a stop line, after the detector detects a vehicle, the direction of the vehicle can be obtained according to the lane where the vehicle is located, and the arrival rate of the corresponding direction of the vehicle is counted. Based on the real-time arrival rate of each flow direction, the two-stage fuzzy controller with similar structure provided by the invention is adopted to control the simulated intersection.
(3) Fuzzy control scheme based on vehicle travel time and path prediction
In the scheme 3, a traffic signal fuzzy control mode based on the predicted arrival rate is adopted, a detector is respectively arranged at the position 300 meters upstream of each entrance stop line, and when a vehicle arrives at the detector, the detector reads vehicle identification information and calls a travel time prediction algorithm and a path prediction algorithm, and the time of the vehicle arriving at an intersection and the running direction of the vehicle are predicted based on the vehicle history travel information. The arrival prediction information of the individual vehicles is corresponding to the signal period and the phase of the intersection to obtain the predicted arrival rate of each phase of the intersection in each period, and the intersection is controlled by using the traffic signal fuzzy control mode provided by the method based on the predicted arrival rate.
The minimum green time, the saturation flow rate, the phase setting and the like of the scheme 2 and the scheme 3 adopt the same parameters as the timing signal control, and the control effect comparison is ensured under the same conditions. Scheme 2 and scheme 3 detector arrangement reference is made to fig. 7a, 7b. Further, the maximum green light time is set to 50s, and the output green light extension time is set to 0 to 40s.
In order to simulate real flat peak and peak flow, the three schemes generate corresponding quantity of traffic flows according to the table 1, the schemes 1 and 2 directly define the number of vehicles in each flow direction, the scheme 3 extracts a vehicle travel time data set and a vehicle path data set of a research intersection according to the preprocessed license plate recognition data, and vehicles with two or more times are input into a simulation environment and are randomly generated according to the corresponding traffic flow.
Under the condition of unified traffic flow input, firstly, aiming at intersections under three schemes of scheme 1, scheme 2 and scheme 3, simulation experiments are respectively carried out under two traffic states of a flat peak period and a peak period. The simulated preheating time is 200s and the effective simulation time is set to 1800s.
And (3) performing simulation experiments on two traffic states, continuously recording the average delay of vehicles at the intersection, finally outputting statistical results of the average delay of the vehicles, the queuing length and the parking times, repeating 10 times of experiments, and calculating the average value of three evaluation indexes obtained by three methods of multiple simulation experiments, wherein the results are shown in Table 2.
Table 2 comparison of signal control effect evaluation indexes for three schemes
The average delay of the method is 30.29 seconds and 36.43 seconds respectively, which are respectively improved by 33.81 percent and 35.34 percent compared with timing signal control, and 22.87 percent and 19.98 percent compared with traditional fuzzy control. In the aspect of queuing length, the method improves the timing signal control by 30.16% and 34.15% respectively in the peak period and the peak period, and improves the timing signal control by 19.89% and 14.59% respectively in the traditional fuzzy control. This shows that the method can reasonably arrange green light extension time by predicting the time from the upstream to the intersection of vehicles, thereby reducing the queuing length. In the aspect of the times of stopping, the method improves by 26.74 percent and 21.98 percent respectively compared with timing signal control and improves by 18.18 percent and 16.47 percent compared with traditional fuzzy control in the peak time and the peak time. The method can effectively reduce the parking times of the vehicle, reduce the secondary parking of the vehicle, and can arrange reasonable green light extension time by predicting the time of the vehicle reaching the intersection, so that the vehicle which should be parked originally and wait for the next period to pass through in the period, and further reduce the parking times of the vehicle. Therefore, the method can effectively improve the passing efficiency of the intersection.
The above examples of the present invention are only for describing the calculation model and calculation flow of the present invention in detail, and are not limiting of the embodiments of the present invention. For ease of description, the present invention names intersections by orientation, but this does not limit the application of the present invention, and this statement applies to all orientation-based descriptions of the present invention. Other variations and modifications of the above description will be apparent to those of ordinary skill in the art, and it is not intended to be exhaustive of all embodiments, all of which are within the scope of the invention.
Claims (10)
1. A traffic signal fuzzy control method considering individual prior data is characterized by comprising the following steps:
step 1, cleaning and analyzing license plate identification data;
step 2, performing personalized travel time prediction of a support vector machine driver based on Bayesian optimization;
step 3, performing Bayesian driver personalized path prediction based on Markov chains;
step 4, fusing and utilizing the historical trip information and the real-time detection information of the vehicle;
and 5, obtaining the phase arrival rate and the turning rate through vehicle travel time and path prediction, and constructing a traffic signal fuzzy control method.
2. The traffic signal fuzzy control method considering individual prior data according to claim 1, wherein in step 1, individual vehicle historical trip data is obtained through a traffic detection technology, a historical trip database is constructed, and the historical trip data is cleaned and analyzed.
3. The traffic signal fuzzy control method considering individual prior data according to claim 1, wherein in step 2, a support vector machine model is adopted to predict the travel time of the driver, and the weather, the time period, the average travel time of the road section and the average travel time characteristic of the driver are utilized to predict the travel time of the driver in the future time period; x is x i To influence the factors of travel time prediction, x j For travel timeThe radial basis function and decision function of the support vector machine travel time prediction model are as follows:
k(x i ,x j )＝exp(-γ||x i -x j || 2 ) (1)
wherein alpha is i For the solution of the dual problem converted by utilizing the Lagrangian multiplier method, gamma is the bandwidth of a kernel function, and b is a bias term;
in the process of predicting the travel time by using the model, searching an optimal parameter combination of the radial basis function bandwidth parameter gamma, the penalty coefficient C and the insensitive coefficient epsilon by adopting a Bayesian optimization method, namely calculating posterior probability by using prior probability and observation data to perform parameter optimization;
the proxy model is as follows:
the mean function is μ (x), the covariance function is k (x, x'), and the Gaussian process is expressed as
f(x)～GP(μ(x),k(x,x′)) (3)
The sampling strategy is used for determining the value of the next set of super parameters, and the used sampling strategy is expected to be improved, and the formula is as follows:
wherein μ (x) is the posterior distribution mean, σ (x) is the covariance, f (x) t+1 ) Is an optimal objective function.
4. The traffic signal fuzzy control method considering individual prior data according to claim 1, wherein in step 3, a bayesian method is utilized to predict a vehicle path, real-time license plate recognition is realized according to a bayonet monitoring device,determining a partial travel path of the vehicle to obtain a vehicle front path T q Predicting the position of the vehicle passing through the next bayonet, converting the problem into calculating a known preamble path T based on a Bayesian model q Under the condition that the vehicle passes through the next intersection l j Probability of (2):
wherein: p (l) j ) For the prior probability, the position of the next bayonet of the vehicle is l j The probability of passing through the bayonet l j A ratio of the number of paths to the amount of historical path data;
in the process of predicting the travel path of the individual vehicle, the travel path is regarded as a Markov chain, the next position of the driver is considered to be related to the current position of the driver, and the primary travel path T= { T i ,L,t j },t k E R is decomposed into T= { T i ,t i+1 }{t i+1 ,t i+2 }L{t j-1 ,t j },t k ∈R
Individual vehicle passing path t= { T i ,t i+1 }{t i+1 ,t i+2 }L{t j-1 ,t j },t k After e R, the probability of going from Bayonet li to Bayonet lj is as follows:
P(T)＝p i,i+1 p i+1,i+2 Lp j-1,j (6)
wherein: p is p ij For two adjacent bayonets l i ，l j The state transition probability of (1) corresponds to the state transition probability of the interface i Driving direction bayonet l j The probability of (1) is that it contains l i -l j The number of historical paths and the inclusion of l i A ratio of the number of historical paths;
constructing a state transition probability matrix M between all adjacent bayonets:
the total path probability between two bayonets is as follows:
wherein: s is the shortest distance between bayonets i, j; k is the detour coefficient
When the vehicle passes through the bayonet j, the front path is T q Probability of (2):
wherein P (T) q ) Is a path probability; p is p c→j The transition probability of the vehicle reaching the next bayonet j from the current position is given; p is p s→j The transition probability of the vehicle from the starting position s to the next bayonet j;
obtaining the vehicle on the known path T according to the formula (1) q Under the condition of (1), the probability of passing through different bayonets at the next moment, and the bayonets with the highest probability are selected as path prediction results.
5. The traffic signal fuzzy control method taking into account individual prior data according to claim 1, wherein in step 4, the traffic detection device extracts the historical travel information of the vehicle, constructs a vehicle personalized travel characteristic database, and comprises the travel mode and the route preference information of the vehicle; comparing the real-time detected vehicle information with a historical database to identify and verify the identity and behavior of the vehicle, and predicting the travel time of the vehicle with high precision by using a travel time prediction model in combination with the current weather, time period, road section average travel time and driver average travel time to predict the travel time of the vehicle from the upstream to the target intersection; predicting the next occurrence position of the vehicle through a vehicle path prediction model according to the travel path track, analyzing the steering selection behavior of the vehicle, obtaining the phase required by the vehicle to pass through an intersection, and predicting the next occurrence position information and occurrence time of the vehicle; the arrival prediction information of the vehicles is corresponding to the signal period and the phase of the intersection, the arrival vehicles of each phase in each period are accumulated and summed, and the arrival vehicles are divided by the period duration to obtain the predicted arrival rate of each phase of the intersection in each period; and after the travel is finished, updating a vehicle personalized travel characteristic database by using the data detected in real time, and analyzing the behavior mode of the individual vehicle.
6. The traffic signal fuzzy control method of claim 1, wherein in step 5, based on the four-phase setting, the traffic demand of each flow direction is predicted in combination with the data of the detector at the upstream of the intersection, and the vehicle travel time is reduced as a target, so as to construct the traffic signal control method based on the two-stage fuzzy controller.
7. The traffic signal fuzzy control method considering individual prior data according to claim 6, wherein in step 5, a signal control system is initialized, and a first-stage fuzzy controller is used for outputting preliminary green lamp delay according to the green lamp phase prediction arrival rate; the first-stage fuzzy controller is a current green light phase traffic intensity judging module, and the module preliminarily outputs green light extension time according to the arrival condition of vehicles in green light phases, and the design steps are as follows:
(1) Input and output of fuzzy controller
Input variables: predicted arrival rate lambda of current green light phase gi
Output variable: green light delay DT 1
(2) Variable domain of discussion and scale factor
The upstream detector is arranged at the starting point of a section of connection between the upstream intersection and the target intersection, and the downstream detector is arranged at the stop line of the entrance road of the target intersection; according to the actual traffic flow condition of the research area and related research, the current predicted arrival rate lambda of the green light phase gi Is set to [0,0.3 ]]The scale factor takes 1, and the fuzzy universe is [0,0.3]The method comprises the steps of carrying out a first treatment on the surface of the Green light delay DT 1 The basic domain is set to [0,30]The scale factor takes 1, and the fuzzy universe is [0,30]
(3) Fuzzy subset of input-output variables
Predicted arrival rate lambda of current green light phase gi Dividing into 7 fuzzy subsets, respectively { very low, medium, high, very high }, denoted { VL, L, RL, M, RH, H, VH }; all input and output variables of the fuzzy controller are divided into 7 fuzzy subsets
(4) Membership function
Selecting a triangle membership function
(5) Control rules
The first-stage fuzzy controller is a single-input single-output fuzzy controller, and the traffic intensity of the current green light phase, namely the predicted arrival rate lambda of the vehicle gi The larger the extension time corresponding to the green light phase is, the longer the extension time is; conversely, the predicted arrival rate lambda of the vehicle at the current green light phase gi The smaller the extension time corresponding to the green light phase is, the shorter the extension time is; 7 corresponding rules are formulated according to fuzzy subset division of the input and output variables;
(6) Deblurring
The gravity center method is used for resolving ambiguity, all functions of the primary fuzzy controller are achieved through programming, and input variables are set: predicted arrival rate lambda of current green light phase gi And obtaining an output variable after fuzzification, fuzzy reasoning and fuzzy operation: green light delay DT 1 Is a value of (a).
8. The traffic signal fuzzy control method considering individual prior data according to claim 6, wherein in step 5, after the target phase is converted into a red light, the target phase enters a second-stage fuzzy controller, the arrival rate is predicted according to the red light phase, and green light delay compensation is output by using the second-stage fuzzy controller; the second-stage fuzzy controller is a current red-light phase traffic intensity judging module, the module judges the traffic intensity of the red-light phase from the current phase to the initial green-light phase through fuzzy reasoning according to the green-light extension time output by the first-stage fuzzy controller according to the arrival condition of the current three red-light phase vehicles, and outputs green-light delay compensation time, and the design steps are as follows:
(1) Input and output of fuzzy controller
Input variables: green light delay DT 1 Maximum predicted arrival rate lambda of current red light phase ri
Output variable: green light delay compensation DT 2
(2) Variable domain of discussion and scale factor
Maximum predicted arrival rate lambda of current red light phase ri Is set to [0,0.3 ]]The scale factor takes 1, and the fuzzy universe is [0,0.3]The method comprises the steps of carrying out a first treatment on the surface of the Green light delay DT 1 The basic domain is set to [0,30]Due to the delay compensation time DT 2 Cannot exceed the green light delay DT 1 So green light delay DT 1 The scale factor of (2) is taken as 2, and the fuzzy discourse domain is [0,60 ]]The method comprises the steps of carrying out a first treatment on the surface of the Green light delay compensation DT 2 The basic domain is set to [0,30]The scale factor takes 1, and the fuzzy universe is [0,30]
(3) Fuzzy subset of input-output variables
Maximum predicted arrival rate lambda of current red light phase ri Green light delay DT 1 Green light delay compensation DT 2 Are all divided into 7 fuzzy subsets
(4) Membership function
Consistent with the first-level fuzzy controller, a triangle membership function is selected
(5) Control rules
The second-stage fuzzy controller is a multi-input single-output fuzzy controller, and predicts the arrival rate lambda ri The larger the green light delay DT 1 The longer the delay compensation time DT 2 The longer; conversely, the predicted arrival rate lambda ri Smaller green light delay DT 1 The shorter the delay compensation time DT 2 The shorter; according to fuzzy subset division of input and output variables, 49 corresponding rules are formulated
(6) Deblurring
Adopting a gravity center method to deblur, and obtaining a deblurred output curved surface through programming; for lambda ri And DT (DT) 1 Can find the unique correspondence in the defuzzified curved surface and along with lambda ri And DT (DT) 1 Is reduced, green light delay compensated DT 2 And also decreases with lambda ri And DT (DT) 1 Is increased, green light delay compensation DT 2 Also along withAn increase in size;
and outputting a green lamp delay according to the actual situation of the intersection, and executing the next phase.
9. A system for implementing the traffic signal fuzzy control method taking into account individual prior data as defined in any one of claims 1 to 8, characterized in that the system comprises a computer module containing the traffic signal fuzzy control method taking into account individual prior data.
10. The system of claim 9, wherein the computer module comprises a first level fuzzy controller, a second level fuzzy controller, a decision module, and a phase switching module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310815824.3A CN116805449A (en) | 2023-07-05 | 2023-07-05 | Traffic signal fuzzy control method and system considering individual priori data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310815824.3A CN116805449A (en) | 2023-07-05 | 2023-07-05 | Traffic signal fuzzy control method and system considering individual priori data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116805449A true CN116805449A (en) | 2023-09-26 |
Family
ID=88079467
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310815824.3A Pending CN116805449A (en) | 2023-07-05 | 2023-07-05 | Traffic signal fuzzy control method and system considering individual priori data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116805449A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117804490A (en) * | 2024-02-28 | 2024-04-02 | 四川交通职业技术学院 | Comprehensive planning method and device for vehicle running route |
-
2023
- 2023-07-05 CN CN202310815824.3A patent/CN116805449A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117804490A (en) * | 2024-02-28 | 2024-04-02 | 四川交通职业技术学院 | Comprehensive planning method and device for vehicle running route |
| CN117804490B (en) * | 2024-02-28 | 2024-05-17 | 四川交通职业技术学院 | Comprehensive planning method and device for vehicle running route |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2610836B1 (en) | Device and method for the on-line prediction of the driving cycle in an automotive vehicle | |
| CN112629533B (en) | Fine path planning method based on road network rasterization road traffic prediction | |
| CN112700072A (en) | Traffic condition prediction method, electronic device, and storage medium | |
| CN107067759B (en) | ACP-based parallel traffic signal lamp real-time control method | |
| CN114202120B (en) | Urban traffic journey time prediction method for multi-source heterogeneous data | |
| CN104464310A (en) | Signal collaborative optimization control method and system of multiple intersections of urban region | |
| CN112863182B (en) | Cross-modal data prediction method based on transfer learning | |
| CN114495499B (en) | Multi-target intelligent internet vehicle cooperative optimization control method | |
| CN116805449A (en) | Traffic signal fuzzy control method and system considering individual priori data | |
| CN110363984A (en) | Traffic flow forecasting method and equipment | |
| CN115409256A (en) | Route recommendation method for congestion area avoidance based on travel time prediction | |
| CN115798199A (en) | Vehicle path dynamic self-adaptive planning method, system, computer equipment and storage medium | |
| CN116311948B (en) | Vehicle path planning method based on traffic flow speed prediction and signal lamp state | |
| CN113112823A (en) | Urban road network traffic signal control method based on MPC | |
| CN113516277A (en) | Network connection intelligent traffic path planning method based on dynamic pricing of road network | |
| Zheng et al. | Soup: spatial-temporal demand forecasting and competitive supply in transportation | |
| CN115862322A (en) | Vehicle variable speed limit control optimization method, system, medium and equipment | |
| Mohammadi et al. | User throughput optimization for signalized intersection in a connected vehicle environment | |
| Du et al. | Dynamic capacity estimation of mixed traffic flows with application in adaptive traffic signal control | |
| Hu et al. | SOUP: A fleet management system for passenger demand prediction and competitive taxi supply | |
| Bao et al. | A personalized route search method based on joint driving and vehicular behavior recognition | |
| Reddy et al. | Survey on traffic flow prediction for intelligent transportation system using machine learning | |
| CN115376308B (en) | Prediction method for automobile running time | |
| Fehnker | Application of evolutionary algorithms to analyze criticality in urban traffic scenarios | |
| Yumaganov et al. | Cooperative Application of Vehicular Traffic Rerouting Method and Adaptive Traffic Signal Control Method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |