CN109448369B - Real-time operation risk calculation method for expressway - Google Patents
Real-time operation risk calculation method for expressway Download PDFInfo
- Publication number
- CN109448369B CN109448369B CN201811260430.1A CN201811260430A CN109448369B CN 109448369 B CN109448369 B CN 109448369B CN 201811260430 A CN201811260430 A CN 201811260430A CN 109448369 B CN109448369 B CN 109448369B
- Authority
- CN
- China
- Prior art keywords
- risk
- factors
- entropy
- real
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
Abstract
The invention relates to a real-time running risk calculation method for an expressway, which adopts a multi-factor fusion entropy model to analyze and calculate real-time running conditions of expressway traffic flow and road by environment. The method is based on the principle of an entropy model, on the basis of acquiring real-time traffic flow data and real-time environment change data, preprocessing the acquired data, and fusing traffic flow risk factors and external environment factors according to the real-time running condition of the expressway by using methods such as correlation analysis, a random forest model, normalization processing and the like to establish a risk entropy model capable of analyzing the real-time running condition of the expressway. The model can identify, evaluate and analyze real-time operation risks of the highway, can provide basis for real-time risk management of the highway, and managers can carry out risk treatment and prevention on the basis, so that economic and property losses caused by high-risk conditions are avoided or reduced.
Description
Technical Field
The invention relates to the technical field of highway traffic, in particular to a real-time operation risk calculation method for a highway.
Background
With the increasing traffic demand, the highway is faced with the problems of traffic jam and traffic safety, and the higher the driving risk of the road, the more seriously the vehicle driving safety is affected. In practice, traffic safety and traffic accidents are two mutually opposite concepts, and the road is not subjected to accidents and does not mean complete safety, and the risk of the occurrence of the traffic accidents still exists at the moment. Therefore, in order to meet the basic characteristics of high speed, high efficiency, safety and comfort of the highway, the risk problem needs to be properly analyzed, and targeted risk management is carried out according to the risk state of road operation.
In reality, factors influencing traffic safety are complex, including differences of road infrastructure and control, and also including influences of weather factors and traffic accidents, but in the aspect of real-time running risks of the expressway, the factors influencing time-varying risks are mainly traffic flow factors and environmental factors. At present, risk evaluation methods adopted in the research of highway operation risks at home and abroad comprise an analytic hierarchy process, a Monte Carlo simulation technology, a sensitivity analysis method, a fuzzy analysis method, a statistical probability method, an influence graph analysis method and the like. The methods approximately reflect the existing level of engineering risk analysis technology, but lack certain real-time reactivity in practical application, and have weak real-time feedback on road running conditions. In view of the existing research results, a complete risk assessment method which is oriented to macroscopic traffic flow operation and can reflect the real-time operation condition of the road is not available, so that risk decision support is provided for a road manager.
Disclosure of Invention
The invention aims to provide a real-time running risk calculation method for an expressway, which is used for calculating a road traffic running risk value from the macroscopic view of traffic flow running, integrating traffic flow factors, weather factors, accident factors and the like and evaluating the risk condition of the running state of the expressway.
The technical scheme adopted by the invention is as follows:
the real-time operation risk calculation method for the expressway is characterized by comprising the following steps:
the method adopts an entropy model of multi-factor fusion to analyze and calculate the real-time running conditions of the traffic flow and the environment of the highway to the road.
The method specifically comprises the following steps:
step 1: classifying factors influencing real-time operation risks of the expressway into two categories, namely traffic flow risk factors and external environment risk factors;
step 2: determining equipment and a method used for acquiring each factor, preprocessing data, and removing and repairing abnormal data;
and step 3: dividing the determined risk factors into positive and negative entropy factors, and performing positive and negative entropy feature screening and entropy weight determination by using feature engineering on the basis of normalization processing according to the obtained factors;
under the assistance of historical accident data analysis and real-time data state analysis, combining the results of normalization processing, and based on the descriptive indexes of all factors, performing feature screening and weight determination by using a feature engineering method to finally obtain the weight values of all factors;
and 4, step 4: and (3) establishing a model for calculating according to the attribute of the influence factors of the entropy model and the weight determination result and the weight analysis result in the step (3) aiming at the attribute difference of the positive entropy and the negative entropy of the highway operation risk, wherein the sum of the positive entropy and the negative entropy is the highway operation risk entropy.
In the step 1, the traffic flow risk factors comprise traffic volume, speed difference, head space, crowding degree and large-scale vehicle proportion, and the external environment risk factors comprise road traffic accident factors and weather factors.
In step 2, for the traffic flow risk factors, selecting a coil detector and a microwave detector to obtain various traffic flow information in real time; for external environment risk factors, accident information, namely abnormal weather information, is acquired from an accident information publishing platform, and weather information is acquired from a roadside road weather station in real time.
In the positive-negative entropy factor division, traffic volume, speed difference, crowding degree, large-scale vehicle proportion, weather and accidents are positive entropy factors, and the distance between the two heads is a negative entropy factor;
in step 3, the weight determination result of each traffic flow factor and external environment factor is as follows:
in step 4, for n evaluation levels, mAPositive entropy factor, mBRisk positive-negative entropy value model of highway operation system under influence of negative entropy factors and overall risk entropy value calculation of systemThe following were used:
in the formula: a is a positive entropy value; b is a negative entropy value; m is an expressway operation risk entropy value; lambda [ alpha ]iIs the weight of the i index; n is the common evaluation grade of each factor; e.g. of the typeiEntropy values of the evaluation indexes are obtained; giIs the difference coefficient of each factor; p is a radical ofiEntropy specific gravity of each factor; n is the importance evaluation value of the ith factor.
The invention has the following advantages:
the method is based on the principle of an entropy model, on the basis of acquiring real-time traffic flow data and real-time environment change data, preprocessing the acquired real-time data, fusing traffic flow risk factors and external environment factors according to the real-time running condition of the expressway by using methods such as correlation analysis, a random forest model, normalization processing and the like, establishing the entropy model capable of analyzing the real-time running condition of the expressway, measuring the real-time running risk of the expressway by using the risk entropy value, and judging the running state of each road section of the expressway from the change state of the risk entropy value. The model digitalizes the running state of the traffic flow, can visually analyze the running state of the road from the perspective of safety risk, can identify, evaluate and analyze the real-time running risk of the highway, can provide basis for the real-time risk management of the highway, and can prevent or reduce the economic and property loss caused by high-risk conditions by observing the change of the risk entropy value and taking corresponding risk treatment and precautionary measures by managers.
Drawings
FIG. 1 is a flow chart of data preprocessing.
Fig. 2 is an importance ranking diagram of traffic risk factors.
Fig. 3 is a flow chart of the concept of the highway operation risk entropy model.
Fig. 4 is a diagram illustrating a correlation analysis between a traffic flow factor and an external environment factor.
Fig. 5 is a time-varying risk graph (west high new) for the highway traveling in the upstream direction.
Fig. 6 is a time-varying risk graph (qu jiang west) for the highway traveling in the upstream direction.
Fig. 7 is a time-varying risk graph (qujianto) of the highway traveling in the upward direction.
Fig. 8 is a down-highway operational risk time-varying graph (west high new).
Fig. 9 is a time-varying risk graph (qu jiang west) for the highway traveling in the down direction.
Fig. 10 is a time-varying risk graph (qujianto) for the highway traveling in the downward direction.
Fig. 11 is a general flow chart of the present invention.
Detailed Description
The present invention will be described in detail with reference to specific embodiments.
The invention relates to a real-time operation risk calculation method for an expressway, which introduces an entropy model into risk analysis, provides a quantifiable road risk assessment method, determines an acquisition method of each factor from expressway operation risk influence factors, quantifies each risk factor, establishes a positive and negative safety entropy model, and analyzes the time-varying property of the real-time operation risk of the expressway based on the variability of the safety entropy.
The specific implementation process of the invention is as follows:
the method comprises the following steps: determining risk factors
In the risk factor research, various risk factors are classified, various influence factors are summarized, and the risk factors are divided into traffic flow risk factors and environmental risk factors.
1. Traffic risk factor
In selecting an index for evaluating a traffic risk from the viewpoint of traffic safety, it is necessary to select data that comprehensively reflects the flow rate, density, and speed as an evaluation index. When determining the traffic risk factors, the method adopts the random forest model in the python software to divide the influence importance of common traffic flow factors on the operation risk.
According to the invention, traffic risk factors are determined by adopting a random forest method according to the importance of output results according to historical and real-time data. The random forest partitioning sample type is mainly based on an indicator function.
Wherein the decision model is hk(x) Y is the classification result and I (.) is the indicator function.
The sample classification is expected to be as follows:
l(s1,s2,L,sn)＝∑Pilog2(pi)i＝1,2,L,n
Pi＝Si/S
wherein the data set is S, n is the number of classifications of S, CiIs a class number, PiFor any sample to belong to CiProbability of piIs the probability, s, of the sample belonging to the ith subsetiTo be classified into CiNumber of samples above. l(s)1,s2,L,sn) The smaller the value of (a), s1,s2,L,snThe more ordered, the better the classification.
And according to the random forest model, determining different data values for different expressways and different traffic flow running states, and constructing the random forest model by taking the bidirectional six-lane expressways as an example. In a random forest model considering traffic flow factors, 15 numbers are independently constructed, each number represents one traffic factor, namely n is 15. The number of risk factors is unchanged during the analysis. Assuming that the running risk of the expressway has N samples, generating a decision tree training set of a running risk model by a repeated sampling method, wherein the total M input variables are S-M. And (3) randomly selecting M (M is less than M) specific variables by each risk node, and determining the optimal splitting point by using the M variables. During the generation of the decision tree, the value of m remains unchanged. In the random forest model calculation, the decision tree is not pruned, and finally, risk factors in traffic factors are determined according to the importance of each factor.
And meanwhile, classification evaluation is carried out on samples in random forest by adopting a Gini index, and the smaller the Gini index is, the lower the probability that the sample is mistakenly divided is, and the purer the sample set is.
Index of kini
And determining the traffic volume, the speed difference, the distance between the vehicle heads, the congestion degree and the proportion of the large-scale vehicles as traffic risk factors by referring to the random forest analysis result. The analysis result of the importance degree of the traffic risk factor is shown in figure 1.
Traffic volume
The traffic volume is the standard number of vehicles passing through the road section in unit time (pcu). Under the same road environment condition, when the road does not exceed a traffic saturation state, the larger the traffic volume is, the worse the running safety performance of the road is; in a saturated state, the crowding degree is increased, the distance between the vehicle heads is reduced, the running speed is reduced, and the fault-tolerant capability is reduced; in the turbulent state, the unbalance of the traffic flow operation occurs, and the operation safety is rapidly reduced in the state.
Velocity difference
The speed difference in the description refers to the difference of average speeds of different vehicle types. Generally, the driving rule of the expressway is vehicle type division and lane division driving, and the higher the speed difference of different vehicle types is, the greater the running risk of the expressway is, so that when analyzing the running risk of the road, the influence of the speed difference on the running risk of the road needs to be considered.
③ interval between headstock and outside
The distance between the vehicle heads is the spacing distance between the front and rear adjacent vehicle heads in a line of vehicle fleets running in the same direction on a lane. To a certain extent, the vehicle head distance can reflect the traffic flow density condition of a road, the larger the traffic flow density is, the smaller the vehicle head distance is, the smaller the fault-tolerant space left for a driver is, and the more the high risk condition is.
Degree of crowding
The congestion degree is equivalent medium-sized vehicle average traffic volume q 'in a certain unit time period of the road section'mEquivalent traffic adaptive quantity q of medium-sized vehicles on same daymThe ratio of (a) to (b). The degree of congestion can reflect traffic volume and density of traffic flow. Traffic adaptive volumeOn the same premise, the higher the congestion degree is, the higher the traffic volume is, the higher the traffic density is, and the higher the running risk of the expressway is.
Proportion of large-scale vehicle
According to the classification of the representative vehicle types of the road vehicles in China, the medium-sized vehicle, the large-sized vehicle and the vehicle train are used as the calculation vehicle types of the large-sized vehicle proportion. The proportion of the large vehicles indicates the number C of the large vehicles in the statistical unit timeTThe ratio of the total traffic volume C. The statistics of the accident information of the expressway shows that the severity of the accident of the large-sized vehicle is obviously higher than that of the medium-sized vehicle and the small-sized vehicle, and the accident risk and probability of the accident are far higher than those of the medium-sized vehicle and the small-sized vehicle, so that the proportion of the large-sized vehicle is taken as a risk factor. The formula for calculating the proportion of the large-scale vehicle is
PT＝CT/C
2. Environmental risk factor
Environmental risk factors include both road traffic accidents and weather.
Traffic accident
The severity of the accident can be divided according to the grade of the road transportation safety accident, and for the road section i, if n accidents happen to the road section, the accident assessment index of the road section i
In the formula, anThe accident grade influence coefficient of the nth accident is shown.
The table below "statistics of different accident-level influence coefficients" shows different numerical values corresponding to different accident levels.
| Grade of accident | Coefficient of influence |
| Nothing elseTherefore, it is | 0 |
| Minor accident | 0.2 |
| General accident | 0.5 |
| Major accident | 0.8 |
| |
1 |
In the formula, anThe accident grade influence coefficient of the nth accident is shown.
Weather factor-
The influence degree of different weather conditions on traffic safety is different, and when the weather is fine, the influence of the weather conditions on the traffic safety is small; when severe weather such as rainstorm, strong wind and the like occurs, the number of traffic accidents accounts for 5% -10% of the total number of traffic accidents. And calibrating the influence of different weather early warning levels on the traffic state. The following table shows the influence coefficients of different weather early warning levels on traffic safety:
| weather early warning level | Coefficient of influence |
| Without pre-warning | 0.1 |
| Blue warning | 0.3 |
| Yellow early warning | 0.5 |
| Orange early warning | 0.7 |
| Red early warning | 0.9 |
The factors influencing the real-time running risk of the highway can be increased and decreased in time according to the requirement in practical application and by referring to the analysis result of the importance degree.
Step two: acquisition and data preprocessing for determining risk factors
And for the traffic flow risk factors, a traffic flow detector is adopted to obtain the traffic volume, the speed difference and the distance between the two heads of the road section to be analyzed in real time, and the congestion degree and the proportion of the large-scale vehicles are calculated based on the obtained traffic volume information so as to obtain all the traffic flow factors.
In the method described in this patent, the units of calculation for each traffic risk factor are as follows:
traffic volume: in unit time, the number of actual traffic participants passing through a certain section of the highway needs to be converted into a standard passenger car. The unit time described in this patent is 15 minutes, and therefore the unit of traffic volume is pcu/15 min.
Speed difference: in unit time, by the speed difference between small cars and large cars on a certain section of a highway, the cars are classified according to representative highway models, the small cars comprise small coaches (the number of seats is less than or equal to 19, and the number of trucks with the carrying quality is less than or equal to 2 t) and medium cars (the number of seats is more than 19, and the number of 2t is less than or equal to 7 t), and the large cars comprise large cars (the number of 7t is less than or equal to 20t) and motor trains (the carrying quality is more than 20 t). The unit time described in this patent is 15 minutes, so the speed difference is calculated every 15 minutes, in km/h.
The distance between the car heads: in a platoon fleet that the syntropy was gone on a lane, the interval distance between two adjacent locomotive heads around, in this patent, use 15 minutes vehicle to pass through the locomotive interval mean value of a certain section as the judgement index, the unit is m.
Degree of crowding: equivalent medium-sized vehicle average traffic volume q 'passing through a certain cross section in unit time'mEquivalent traffic adaptive quantity q of medium-sized vehicles on same daymThe ratio of (a) to (b). The unit time described in this patent is 15 minutes.
The proportion of the large-scale vehicle is as follows: the ratio of the number of large vehicles passing through a section of the highway to the number of all vehicles passing through the section in unit time. In contrast to the classification of representative types of road automobiles, the small-sized automobiles comprise small passenger automobiles (passenger automobiles with seats less than or equal to 19 seats and trucks with load mass less than or equal to 2 t), the large-sized automobiles comprise medium-sized automobiles (passenger automobiles with seats greater than 19 seats and trucks with load mass less than or equal to 7t, 2 t), large-sized automobiles (trucks with load mass less than or equal to 20t, 7 t) and automobile trains (with load mass greater than or equal to 20 t). The unit time stated in this patent is 15 minutes in%.
For environmental risk factors, road accident factors are historical abnormal traffic events of a risk analysis road section, and can be acquired from an accident information publishing platform, and in the factor analysis, the road traffic event alarm information data is preprocessed, abnormal data is removed and repaired, and accident occurrence information of the road section to be evaluated, including accident occurrence places, accident grades, accident reasons, occurrence time and the like, is acquired; for weather factors, weather information can be acquired in real time by depending on a road weather station, and whether the weather factors belong to weather early warning levels or not is judged.
The data collected by the method has partial invalid data due to the reasons of equipment precision, short-time external environment interference and the like, the data needs to be preprocessed before model calculation, operations such as cleaning, removing, complementing and the like are carried out, and the preprocessed data can be used for specific calculation of the safety entropy model.
And for null data, establishing a regression equation, and filling and complementing by adopting a calculation value of the regression equation.
When abnormal data is eliminated, samplingThe method of removing by factors is used for firstly removing abnormal data under each factor. A 1 × 96 matrix N is formed for the data acquisition situation of each factor in units of 15 minutes with a day periodiCorresponding to the total number of influencing factors in step 1, 7 matrices are formed. When the matrix is established for different roads, the time unit of the matrix element can be adjusted according to the environmental characteristics of the road section to be analyzed.
For a matrix N according to the Grabbs criterioniIf the individual elements deviate far from the average, the data can be determined to be anomalous. The specific judgment process of the abnormal data is as follows:
arranging the elements of each matrix in the order from small to large, and calculating the average value of each matrix
Standard deviation S, ratio G of residual error of each numerical value to standard deviationi。
Will calculate the value GiAnd a Grabbs table threshold value Gp(n) comparing if GiGreater than GpAnd (n), the numerical value is an abnormal numerical value and needs to be removed. If data of one factor in a certain period is rejected, data of other factors in the period also need to be rejected at the same time.
The data after the processing of each factor corresponds to a time period, and a matrix N including 7 × 96 risk analysis elements is formed.
Step three: determining entropy weights
The process of determining the entropy weight is the process of performing correlation analysis and normalization processing on the risk factors. With the aid of the accident data analysis, the result of the data analysis can be used as a safety entropy weight. And calibrating the weights of the traffic flow factors and the external environment factors by adopting a data analysis method, namely a characteristic screening algorithm in characteristic engineering.
The relevance analysis for determining the weight is to analyze the risk factors determined above, measure the relevance degree among the factors and further convert the relevance degree into the weight of each factor. In specific calculation, the influence of each factor on a high risk condition can be analyzed by using a panadas.
According to the correlation analysis result, the traffic volume, the speed difference, the crowding degree, the proportion of the large-scale vehicles and the weather are positive correlation factors, and the distance between the two vehicles is negative correlation factors. Since the output result is the correlation among all factors, the correlation of each factor to the operation risk of the expressway needs to be converted.
Degree of influence of accident factors on operational risk
u22＝u2-u21
Degree of influence of various traffic flow factors on highway operation risk
In the formula: alpha is alpha1nRepresenting the influence degree of each traffic flow factor in the correlation analysis diagram on the high risk condition; u. of1、u2And respectively representing the influence degrees of traffic flow factors and environmental factors on the running risk of the expressway in the first-level indexes.
And calculating the influence degree of each factor on the running risk of the expressway, wherein the result is shown in the following table and a traffic and environmental factor weight value table.
| Risk factors | Traffic volume | Difference in velocity | Head space | Degree of congestion | Large scale vehicle ratio | Weather (weather) | Accident |
| Coefficient of correlation | 0.058 | 0.044 | -0.03 | 0.158 | 0.307 | 0.239 | 0.164 |
Step four: establishing a safety entropy model:
the highway operation risk is the result of multi-factor coupling, and the influence of various factors on the operation risk needs to be comprehensively considered when a risk model is established. When risk conditions are analyzed, the safety entropy model is introduced into safety risk assessment, influences of traffic flow factors and external environment factors are quantitatively analyzed, the concept of risk entropy is provided, and the risk entropy value is used for evaluating the operation risk of the expressway. The larger the entropy, the greater the risk the system is exposed to.
The specific idea of establishing the safety entropy model is as follows: on the basis of traffic flow factors and external environment factor analysis, all the factors are divided into positive and negative entropy factors according to the influence factor attributes of the entropy model, correlation analysis is carried out on the positive and negative entropy factors according to the difference of the positive and negative entropy factor attributes, the weight of each factor is divided, and finally the weight corresponds to the positive and negative entropy factors and the running risk value of the expressway is output. The establishment idea of the safety entropy model is shown in FIG. 2.
The specific establishment idea of the safety entropy model is as follows:
when calculating the running risk of the expressway, the running condition of the road is regarded as a system influenced by different factors. In general, the system entropy values affected by different factors are
In the formula, i represents the serial number of each subsystem, and the value of i is 1-4; j represents a factor contained in the subsystem; siEntropy values (bit) representing the subsystems, positive entropy values being generated by the driver psychological subsystem and the driver physiological subsystem, and negative entropy values being generated by the vehicle subsystem and the environmental subsystem; kiRepresenting the weight of each subsystem; p is a radical ofjRepresents the weight of the entropy value of each influencing factor,
corresponding to the positive and negative attributes in the risk factor correlation analysis, the factors influencing the entropy value of the system can be divided into positive entropy factors and negative entropy factors, which correspond to +/-in the formula. For a highway operation system, the distance between the vehicle heads belongs to a negative entropy factor, and other factors belong to a positive entropy factor. Therefore, it is necessary to establish n evaluation levels, m, from the viewpoint of multi-factor system analysisAPositive entropy factor, mBAnd (3) a road operation system risk entropy value model under the influence of negative entropy factors.
In the formula: a is a positive entropy value; b is a negative entropy value; lambda [ alpha ]iIs the weight of the i index; n is the common evaluation grade of each factor; e.g. of the typeiEntropy values of the evaluation indexes are obtained; giIs the difference coefficient of each factor; p is a radical ofiEntropy specific gravity of each factor; n is the importance evaluation value of the ith factor.
The unit time of the model calculation can be adjusted according to the requirements of practical application.
The following is a practical application based on the above calculation:
the example application of the invention is explained by taking the city-around highway of the city of xi' an as a typical road section of the bidirectional six-lane highway. The west safe city-surrounding high speed is a totally closed city highway with full interchange and access control, besides the function of a common highway, the west safe city-surrounding high speed also simultaneously undertakes the transportation task of a west safe city loop, namely a three-loop main road, the traffic flow is large, the time variation of macroscopic traffic flow is obvious, in addition, a traffic detector is arranged on the west safe city-surrounding high speed, the real-time information of traffic volume, speed difference, distance between the car heads, crowdedness and the like can be collected, and the historical accident information is recorded. And carrying out operation risk analysis on the test piece according to the theoretical framework.
1. Determining operational risk analysis road segment
And analyzing the uplink and downlink operation risks of the high-speed south segment around the city. The specific road section is from west, high and new to the east of the Qujiang overpass, and the data acquisition nodes of the related sections are distributed as shown in the following table.
| Serial number | Position of | Data type | Serial number | Position of | |
|
| 1 | West-Gaoxin | Uplink is carried out | 2 | West-Gaoxin | Downstream | |
| 3 | Qujiang grade separation east | Uplink is carried out | 4 | Qujiang grade | Downstream | |
| 5 | Qujiang grade separation east | Section information | 6 | Qujiang overpass west | Uplink is carried out | |
| 7 | Qujiang overpass west | Downstream | 8 | Qujiang overpass west | Section ofInformation |
2. Risk factor analysis
The real-time traffic flow data is sorted, traffic volume, large and small vehicle speed difference, vehicle head distance, crowding degree and weather data of every 15 minutes are obtained, historical accident information of all sections of the high-speed south section of the city surrounded by xi 'an is gathered through an accident information issuing platform, historical accident number, accident reasons and the like of all road sections to be analyzed are obtained, real-time running risks of the high-speed south section of the city surrounded by xi' an are analyzed, running risk values of three sections in the up and down directions for 24 hours are obtained, and corresponding risk time-varying graphs are drawn. The risk time-varying diagrams in the up and down directions are shown in fig. 5-10.
According to the risk time-varying graph of the high-speed south segment of the city surrounded by the xi 'an city, the method provided by the description can reflect the change situation of the running risk of the highway in real time according to the time-varying property of traffic flow running, and the running risk presents a fluctuation state for the high speed of the city surrounded by the xi' an city; meanwhile, the running risks at night and in the day are different to a certain extent, roads are in a high-risk state in most of the night, the running state in the day belongs to a low risk, and the running risk at night is obviously higher than that in the day.
The invention is not limited to the examples, and any equivalent changes to the technical solution of the invention by a person skilled in the art after reading the description of the invention are covered by the claims of the invention.
Claims (5)
1. The real-time operation risk calculation method for the expressway is characterized by comprising the following steps:
the method adopts an entropy model of multi-factor fusion to analyze and calculate the real-time running conditions of the traffic flow and the environment of the highway to the road;
the method specifically comprises the following steps:
step 1: classifying factors influencing real-time operation risks of the expressway, and dividing influence importance of traffic flow factors into two types, namely traffic flow risk factors and external environment risk factors, by adopting a random forest model;
step 2: determining equipment and a method used for acquiring each factor, preprocessing data, and removing and repairing abnormal data;
for null value data, establishing a regression equation, and filling and complementing by adopting a calculated value of the regression equation; for abnormal data, a method of removing by factors is adopted;
and step 3: dividing the determined risk factors into positive and negative entropy factors, and performing positive and negative entropy feature screening and entropy weight determination by using feature engineering on the basis of normalization processing according to the obtained factors;
under the assistance of historical accident data analysis and real-time data state analysis, combining the results of normalization processing, and based on the descriptive indexes of all factors, performing feature screening and weight determination by using a feature engineering method to finally obtain the weight values of all factors;
and 4, step 4: and (3) establishing a model for calculating according to the attribute of the influence factors of the entropy model and the weight determination result and the weight analysis result in the step (3) aiming at the attribute difference of the positive entropy and the negative entropy of the highway operation risk, wherein the sum of the positive entropy and the negative entropy is the highway operation risk entropy.
2. The real-time highway operational risk calculation method according to claim 1, wherein:
in the step 1, the traffic flow risk factors comprise traffic volume, speed difference, head space, crowding degree and large-scale vehicle proportion, and the external environment risk factors comprise road traffic accident factors and weather factors.
3. The real-time highway operational risk calculation method according to claim 2, wherein:
in step 2, for the traffic flow risk factors, selecting a coil detector and a microwave detector to obtain various traffic flow information in real time; for external environment risk factors, accident information, namely abnormal weather information, is acquired from an accident information publishing platform, and weather information is acquired from a roadside road weather station in real time.
4. The real-time highway operational risk calculation method according to claim 3, wherein:
in the positive-negative entropy factor division, traffic volume, speed difference, crowding degree, large-scale vehicle proportion, weather and accidents are positive entropy factors, and the distance between the two heads is a negative entropy factor;
in step 3, the weight determination result of each traffic flow factor and external environment factor is as follows:
5. the real-time highway operational risk calculation method according to claim 4, wherein:
in step 4, for n evaluation levels, mAPositive entropy factor, mBThe risk positive and negative entropy value model of the highway operation system and the overall risk entropy value of the system under the influence of the negative entropy factors are calculated as follows:
in the formula: a is a positive entropy value; b is a negative entropy value; m is an expressway operation risk entropy value; lambda [ alpha ]iIs the weight of the i index; n is the common evaluation grade of each factor; e.g. of the typeiEntropy values of the evaluation indexes are obtained; giIs the difference coefficient of each factor; p is a radical ofiEntropy specific gravity of each factor; n is the importance evaluation value of the ith factor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811260430.1A CN109448369B (en) | 2018-10-26 | 2018-10-26 | Real-time operation risk calculation method for expressway |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811260430.1A CN109448369B (en) | 2018-10-26 | 2018-10-26 | Real-time operation risk calculation method for expressway |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109448369A CN109448369A (en) | 2019-03-08 |
| CN109448369B true CN109448369B (en) | 2021-08-03 |
Family
ID=65548674
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811260430.1A Active CN109448369B (en) | 2018-10-26 | 2018-10-26 | Real-time operation risk calculation method for expressway |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109448369B (en) |
Families Citing this family (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109919511B (en) * | 2019-03-20 | 2021-02-26 | 西南交通大学 | Existing railway line shape evaluation method and system |
| CN109902971B (en) * | 2019-03-26 | 2023-05-23 | 南通大学 | Highway road side safety risk evaluation and classification method |
| CN110085028A (en) * | 2019-05-05 | 2019-08-02 | 同济大学 | A kind of traffic safety risk stratification management-control method under mist weather condition |
| CN110491154B (en) * | 2019-07-23 | 2021-04-30 | 同济大学 | Suggested vehicle speed making method based on safety risk and distance |
| CN110910641A (en) * | 2019-12-02 | 2020-03-24 | 北京航空航天大学 | Neural network-based real-time risk assessment method for highway traffic operation |
| CN113125042A (en) * | 2019-12-30 | 2021-07-16 | 四川浩诺优科技有限公司 | Intelligent expressway temperature measuring method |
| CN111210165B (en) * | 2020-01-21 | 2023-04-07 | 哈尔滨工业大学 | Vehicle operation risk assessment system based on risk conduction coupling |
| CN111680897A (en) * | 2020-05-27 | 2020-09-18 | 公安部道路交通安全研究中心 | Service management and control method, device and system and computer readable storage medium |
| CN112201038B (en) * | 2020-09-28 | 2021-09-03 | 同济大学 | Road network risk assessment method based on risk of bad driving behavior of single vehicle |
| CN112504348B (en) * | 2020-12-11 | 2023-07-25 | 厦门汇利伟业科技有限公司 | Object state display method and system integrating environmental factors |
| CN112700138A (en) * | 2020-12-31 | 2021-04-23 | 京东数字科技控股股份有限公司 | Method, device and system for road traffic risk management |
| CN113257024B (en) * | 2021-04-29 | 2023-09-12 | 中汽研汽车检验中心(广州)有限公司 | V2I-based rear-end collision prevention early warning method and system for expressway rain and fog environment |
| CN114005274B (en) * | 2021-10-21 | 2023-03-03 | 浙江综合交通大数据开发有限公司 | Expressway toll station graded linkage management and control method based on equivalent weight of passenger car |
| CN114241753B (en) | 2021-12-03 | 2022-11-01 | 东南大学 | Road safety evaluation method and system based on multi-dimensional influence factors |
| CN114005297B (en) * | 2021-12-07 | 2023-05-02 | 北京车晓科技有限公司 | Vehicle team coordinated driving method based on Internet of vehicles |
| CN114283584A (en) * | 2021-12-31 | 2022-04-05 | 云控智行(上海)汽车科技有限公司 | Expressway road condition prediction method under intelligent network connection environment and computer readable storage medium |
| CN114495501B (en) * | 2022-01-27 | 2022-11-15 | 东南大学 | Road safety analysis method based on combined geographic autoregressive matching |
| CN114613127B (en) * | 2022-02-10 | 2023-04-07 | 江苏大学 | Driving risk prediction method based on multi-layer multi-dimensional index system |
| CN114613130B (en) * | 2022-02-18 | 2023-05-12 | 北京理工大学 | Driving credibility analysis method in traffic and carrying system |
| CN116778733A (en) * | 2022-11-26 | 2023-09-19 | 武汉广旺科技有限公司 | Highway navigation voice early warning method and system based on big data |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103162982A (en) * | 2013-02-04 | 2013-06-19 | 河海大学 | Diagnostic method of evolution state of damage field of concrete dam |
| CN106056308A (en) * | 2016-06-13 | 2016-10-26 | 宁波工程学院 | Highway tunnel operation environment safety risk automatic judgment method |
| CN106651182A (en) * | 2016-12-25 | 2017-05-10 | 北京工业大学 | Rail passenger flow congestion risk evaluation method based on entropy weight |
| CN107424409A (en) * | 2017-06-15 | 2017-12-01 | 深圳源广安智能科技有限公司 | A kind of urban traffic status evaluation system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9047767B2 (en) * | 2013-09-09 | 2015-06-02 | International Business Machines Corporation | Traffic impact prediction for multiple event planning |
-
2018
- 2018-10-26 CN CN201811260430.1A patent/CN109448369B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103162982A (en) * | 2013-02-04 | 2013-06-19 | 河海大学 | Diagnostic method of evolution state of damage field of concrete dam |
| CN106056308A (en) * | 2016-06-13 | 2016-10-26 | 宁波工程学院 | Highway tunnel operation environment safety risk automatic judgment method |
| CN106651182A (en) * | 2016-12-25 | 2017-05-10 | 北京工业大学 | Rail passenger flow congestion risk evaluation method based on entropy weight |
| CN107424409A (en) * | 2017-06-15 | 2017-12-01 | 深圳源广安智能科技有限公司 | A kind of urban traffic status evaluation system |
Non-Patent Citations (1)
| Title |
|---|
| 智慧城市集成服务企业竞争力评价研究;张苏晶;《中国优秀硕士学位论文全文数据库 经济与管理科学辑》;20180131;J155-49 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109448369A (en) | 2019-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109448369B (en) | Real-time operation risk calculation method for expressway | |
| WO2020244288A1 (en) | Method and apparatus for evaluating truck driving behaviour based on gps trajectory data | |
| CN112508392B (en) | Dynamic evaluation method for traffic conflict risk of hidden danger road section of mountain area double-lane highway | |
| CN112201038B (en) | Road network risk assessment method based on risk of bad driving behavior of single vehicle | |
| CN102136190B (en) | Dispatching management system and method for event emergency response of urban bus passenger transport | |
| CN106127586A (en) | Vehicle insurance rate aid decision-making system under big data age | |
| CN110164132B (en) | Method and system for detecting road traffic abnormity | |
| CN109493566B (en) | Fatigue driving prediction method based on GPS data | |
| CN111242484B (en) | Vehicle risk comprehensive evaluation method based on transition probability | |
| CN114783183A (en) | Monitoring method and system based on traffic situation algorithm | |
| CN109408557B (en) | Traffic accident cause analysis method based on multiple correspondences and K-means clustering | |
| CN108550263A (en) | A kind of traffic accidents genetic analysis method based on fault tree models | |
| Rong et al. | Effects of individual differences on driving behavior and traffic flow characteristics | |
| CN115240431A (en) | Real-time online simulation system and method for traffic flow of highway toll station | |
| CN112382086A (en) | Application method of open road test evaluation system of automatic driving automobile | |
| CN114005274B (en) | Expressway toll station graded linkage management and control method based on equivalent weight of passenger car | |
| CN111563555A (en) | Driver driving behavior analysis method and system | |
| CN112990544B (en) | Traffic accident prediction method for expressway intersection area | |
| CN100481153C (en) | Method for automatically inspecting highway traffic event based on offset minimum binary theory | |
| CN114970943B (en) | Expressway accident type and severity prediction method | |
| CN116597642A (en) | Traffic jam condition prediction method and system | |
| CN115311858A (en) | Urban road section grading control method based on traffic flow toughness | |
| CN109405962B (en) | Road traffic noise frequency spectrum calculation method | |
| Wei et al. | Exploring driver injury severity in single-vehicle crashes under foggy weather and clear weather | |
| CN114299742A (en) | Dynamic recognition and updating recommendation method for speed limit information of expressway |
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 |