CN107392373A - A kind of linear index selection and optimization method based on sensitivity analysis and traffic safety - Google Patents

A kind of linear index selection and optimization method based on sensitivity analysis and traffic safety Download PDF

Info

Publication number
CN107392373A
CN107392373A CN201710592024.4A CN201710592024A CN107392373A CN 107392373 A CN107392373 A CN 107392373A CN 201710592024 A CN201710592024 A CN 201710592024A CN 107392373 A CN107392373 A CN 107392373A
Authority
CN
China
Prior art keywords
mrow
section
msub
accident
linear index
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201710592024.4A
Other languages
Chinese (zh)
Other versions
CN107392373B (en
Inventor
孟祥海
梁心雨
蒋艳辉
张道玉
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Harbin Institute of Technology
Original Assignee
Harbin Institute of Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Harbin Institute of Technology filed Critical Harbin Institute of Technology
Priority to CN201710592024.4A priority Critical patent/CN107392373B/en
Publication of CN107392373A publication Critical patent/CN107392373A/en
Application granted granted Critical
Publication of CN107392373B publication Critical patent/CN107392373B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management

Abstract

The present invention relates to highway route index and accidental rate analysis field, more particularly to the linear index selection and optimization method based on sensitivity analysis and traffic safety, the present invention fails to influence for flat vertical combination section research accident rate to solve existing accidental rate analysis method, and then the shortcomings that causing hair section easy to accident identification not accurate enough, method of the invention includes：For each section of highway, the average every kilometer etesian traffic accident quantity that prediction obtains is calculated；Linear index correction factor is set for each section；Calculate the accident prediction model in section；Linear index is chosen, and the sensitivity coefficient of each linear index is calculated according to the type in section；The minimum linear index of sensitivity coefficient is chosen, calculates the growth factor of accident rate；The high linear indication range of degree of safety is filtered out according to the relation of linear index and accident rate growth factor.The security screening that the present invention is applied to highway differentiates.

Description

A kind of linear index selection and optimization method based on sensitivity analysis and traffic safety
Technical field
The present invention relates to highway route index and accidental rate analysis field, and in particular to one kind is based on sensitivity analysis With the linear index selection and optimization method of traffic safety.
Background technology
In recent years, highway in China construction develops rapidly, and expressway construction achieves achievement highly visible.It is but same with this When, with the fast development of China's highway, traffic accidents rise year by year, traffic accidents death people The obvious increase of number, the security of highway fail effectively to embody.Highway is providing conveniently trip together to us When, traffic safety is also the most important thing that we pursue.
The safety guarantee of highway is a complicated system engineering, wherein, the identification of crucial linear index is traffic The essential step of safety guarantee.
The research of domestic road alignment and accident rate mainly has：Before beam summer of Tongji University, Guo Zhongyin, Fang Shouen application/ Post analysis method have studied road alignment and the statistical relationship of accident rate.Harbin Institute of Technology have studied road conditions and traffic thing Therefore dependency relation, and the countermeasure prevented accident is proposed from highway layout condition.The blue will of Jiaotongguihua Design Institute, Hubei Province It is straight line on Chen Yongsheng, Wang Guang mountain joint study of male, Song Jihong and Beijing University of Technology highway, horizontal curve, longitudinal slope, flat The dependency relation of the linear indexs and accident rate such as vertical cooperation, so as to provide technology branch for the security audit during highway layout Hold.
Tamar Ben-Bassat etc. think that highway layout can significantly affect the driving behavior of driver, geometry linear, road Width of shoulder etc. can significantly affect Vehicle Speed, so as to influence traffic safety；FU Rui etc.] pass through substantial amounts of data point Analysis, it is believed that road plane is linear and vertical alignment has significant impact to traffic accident；Equally, the Matthew of Greece G.Karlafti etc. uses the relation between accident rate such as nonparametric statistical method, analysis road geometry linear, flow, finds Highway geometrical design and road surface types are the most important factors of two influence accident rates.
Road plane it is linear with the relation of accident on, the studying plane such as Zelanian Robin Haynes is linear and accident Between relation, find the ratio between link length and point-to-point transmission air line distance and the size of accumulative corner significantly affect accident per km Rate；American scholar Islam and Seneviratne has found that radius of horizontal curve has significantly to the speed of service by field observation data Influence, in the diverse location of curve, the speed of service is also significantly different, and establishing diverse location on horizontal curve by regression analysis transports The model of scanning frequency degree；Malay Ali Aram by studying the traffic accident data in two-lane highway Horizontal Curve Sections, Influenceed it was found that radius of horizontal curve, Horizontal Curve Sections length, superrelation on curve, length of transition curve, shoulder width etc. are formed on accident Larger, curve section is more dangerous than the linear section of equal length, and this phenomenon starts substantially when radius is less than 1000m, Divide protrusion less than 200m ten in radius.
The content of the invention
It is linear with accident relationship analysis, dividing each index the invention aims to solve existing road plane Analysis is more single, and lacks the shortcomings that being screened to crucial linear index and propose a kind of line based on sensitivity analysis and traffic safety Shape index selection and optimization method.
A kind of linear index selection and optimization method based on sensitivity analysis and traffic safety, including：
Step 1: according to the lineament of targeted cache highway, targeted cache highway is divided into k section；The class in section Type includes Mountainous expressway, hills area highway and Expressway in Plain.
Step 2: for each section, the average every kilometer etesian traffic thing that the link prediction obtains is calculated Therefore quantity.
Step 3: set linear index correction factor for each section；Linear index correction factor includes straight line segment length Spend correction factor βLZ, radius of horizontal curve correction factor βRP, horizontal curve drift angle correction factor βα, longitudinal slope gradient correction factor βiAnd Radius of vertical curve correction factor βRS。
Step 4: according to the type in each section, the average every kilometer etesian traffic that the link prediction is obtained Accident quantity is multiplied by all linear index correction factors in the section, obtains the accident prediction model in the section.
Step 5: choose length of straigh line, radius of horizontal curve, horizontal curve drift angle, the longitudinal slope gradient and radius of vertical curve conduct Linear index, and the sensitivity coefficient of each linear index is calculated according to the type in section respectively, the sensitivity coefficient is used to describe Influence degree of the linear index to accident rate predicted value.
Step 6: choosing the minimum linear index of sensitivity coefficient, the growth of the accident rate when linear index changes is asked Coefficient.
Step 7: the high linear index model of degree of safety is filtered out according to the relation of linear index and accident rate growth factor Enclose.
Beneficial effects of the present invention are：
1st, for flat vertical combination section, when studying its crucial linear index difference value, the situation of change of accident rate, fill up The blank of technical field.
2nd, with reference to the method for sensitivity analysis, propose that the crucial linear index of influence traffic safety screen and excellent The method of change, contribute in highway layout and the improvement of later stage safety condition, it is determined that crucial linear index and being adjusted.
3rd, the method according to the invention analysis can be obtained to draw a conclusion：When Mountainous expressway longitudinal slope is 3%, song is put down Line radius value suggestion should be greater than 1500m；When longitudinal slope is 4%, radius of horizontal curve value suggestion should be greater than 2500m；Longitudinal slope is 5% When, radius of horizontal curve value suggestion should be greater than 6500m.When hills area highway longitudinal slope is 3%, radius of horizontal curve value suggestion It should be greater than 1000m；When longitudinal slope is 4%, radius of horizontal curve value suggestion should be greater than 1500m；When longitudinal slope is 5%, radius of horizontal curve Value suggestion should be greater than 7500m.When Mountainous expressway radius of vertical curve is less than 6000m, the suggestion of radius of horizontal curve value should More than 2000m.When hills area highway radius of vertical curve is less than 17000m, radius of horizontal curve value suggestion should be greater than 1500m.
Brief description of the drawings
Fig. 1 is the flow chart of the embodiment of the present invention.
Embodiment
Embodiment one：Present embodiment based on the linear index of sensitivity analysis and traffic safety selection with it is excellent Change method, it is characterised in that including：
Step 1: according to the lineament of targeted cache highway, targeted cache highway is divided into k section；The class in section Type includes Mountainous expressway, hills area highway and Expressway in Plain.
Step 2: for each section, the average every kilometer etesian traffic thing that the link prediction obtains is calculated Therefore quantity.
Step 3: set linear index correction factor for each section；Linear index correction factor includes straight line segment length Spend correction factor βLZ, radius of horizontal curve correction factor βRP, horizontal curve drift angle correction factor βα, longitudinal slope gradient correction factor βiAnd Radius of vertical curve correction factor βRS。
Step 4: according to the type in each section, the average every kilometer etesian traffic that the link prediction is obtained Accident quantity is multiplied by all linear index correction factors in the section, obtains the accident prediction model in the section.
Step 5: choose length of straigh line, radius of horizontal curve, horizontal curve drift angle, the longitudinal slope gradient and radius of vertical curve conduct Linear index, and the sensitivity coefficient of each linear index is calculated according to the type in section respectively, the sensitivity coefficient is used to describe Influence degree of the linear index to accident rate predicted value.
Step 6: choosing the minimum linear index of sensitivity coefficient, the growth of the accident rate when linear index changes is asked Coefficient.
Step 7: the high linear index model of degree of safety is filtered out according to the relation of linear index and accident rate growth factor Enclose.
Embodiment two：Present embodiment is unlike embodiment one：
In step 2,
When section is mountain ridge highway, average every kilometer obtained by equation below calculating prediction is etesian Traffic accident quantity NB：
NB=1.0 × 10-6×AADT1.411
Wherein AADT represents annual average daily traffic；
When section is hills highway, average every kilometer obtained by equation below calculating prediction is etesian Traffic accident quantity NB：
NB=2.06 × 10-7×AADT1.520
When section is Expressway in Plain, is calculated by equation below and predict average every kilometer obtained annual generation Traffic accident quantity NB：
NB=2.2 × 10-9×AADT2-3.75×10-5×AADT+0.96。
Other steps and parameter are identical with embodiment one.
Embodiment three：Present embodiment is unlike embodiment one or two：
In step 4,
When section is Mountainous expressway, the accident prediction model in the section is：
N=NB×βRP×βLZ×βα×βi×βRS
NB=1.0 × 10-6×AADT1.411
βRP=1248.1RP-1.041+1
βLZ=0.264LZ2-0.587LZ+1.44
βα=-9.25 × 10-6×α3+9.38×10-4×α2-2.38×10-2×α+1.204
βi=0.072i2-0.081i+1.226
βRS=12.54RS-0.46+1
When section is hills area highway, the accident prediction model in the section is：
N=NB×βRP×βLZ×βα×βi×βRS
NB=2.06 × 10-7×AADT1.520
βRP=18755RP-1.51+1
βLZ=0.373LZ2-0.758LZ+1.49
βα=-1.19 × 10-5×α3+1.89×10-3×α2-0.08α+1.93
βi=0.138i2+0.019i+0.99
βRS=11.77RS-0.43+1
When section is Expressway in Plain, the accident prediction model in the section is：
N=NB×βLZ
NB=2.2 × 10-9×AADT2-3.75×10-5×AADT+0.96
βLZ=0.247LZ2-0.883LZ+1.91。
Wherein, LZ is length of straigh line, RP is radius of horizontal curve, α is horizontal curve drift angle, i is the longitudinal slope gradient, RS is perpendicular song Line radius.
Other steps and parameter are identical with embodiment one or two.
Embodiment four：Unlike one of present embodiment and embodiment one to three：
In step 5, the expression formula of sensitivity coefficient is：
Wherein m represents to have carried out section m sampling, xi′Represent the sampled value to the i-th ' secondary sampling of progress of linear index, f (xi′) represent accident rate rate of change corresponding with the sampled value of the i-th ' secondary sampling, min { f (x1),f(x2),…f(xm) represent from Minimum value is taken in accident rate of change corresponding to m times all samplings.
Each sensitivity coefficient △ yi′The sensitivity coefficient to the linear index of some individual event is represented, such as in table 1, can be directed to Length of straigh line LZ calculates sensitivity coefficient：Length of straigh line LZ has carried out altogether 10 samplings (i.e. from 0.50km to 2.75km), Correction factor is to calculate what is determined according to the method for prior art；The accident rate rate of change of each section of sampling is first calculated, with the first row Exemplified by, 0.07 is calculated by (1.08-1.01)/1.01, wherein 1.08 be the correction factor of this sampling, 1.01 be institute There is minimum that of value in correction factor, i.e.,Then all accidents are become Rate add and, then divided by hits, what is obtained is exactly the sensitivity coefficient of the linear index of correspondence, i.e. every numerical value in table 2.
Other steps and parameter are identical with one of embodiment one to three.
Embodiment five：Unlike one of present embodiment and embodiment one to four：
In step 6, the calculation formula of accident rate growth factor is：
Wherein ξi″jFor accident rate growth factor；β1i″For in longitudinal slope gradient correction factor i-th " individual data point, i "=1, 2,3,4……；β2jFor j-th of data point in radius of horizontal curve correction factor, j=1,2,3,4 ....
ξ hereini″jThe numerical value that the jth row i-th in table 3 arranges is can be understood as, example is enumerated with the 2nd row the 1st,Wherein β11The longitudinal slope gradient correction factor when gradient is -6 is represented, longitudinal slope gradient correction factor can Calculated with the formula according to prior art, such as the separate equations in formula 5 to formula 9.β22Represent that radius of horizontal curve is Radius of horizontal curve during 250m.Denominator min (β1i·β2i) represent to each β in table1i·β2iCalculated, then selected Minimum value therein.
<Embodiment>
The flow chart of the present embodiment is as shown in figure 1, specifically include：
(1) accident prediction model is established
The idea about modeling of reference IHSDM accident prediction models, accident rate forecast model under the conditions of foundation ideal is linear, then Plane, vertical alignment correction factor are introduced, the progressively amendment of alignment condition is carried out to basic accident prediction model, and then is obtained Accident prediction model under real road condition and traffic environment.Mountainous terrain, hills area and Expressway in Plain accident are pre- The general structure for surveying model is shown in formula 1.
N in formulai--- average every kilometer, etesian traffic accident quantity, i.e. mould on the predicting unit i that prediction obtains The accident rate (secondary/km) of type output；
NBi--- when predicting unit i it is linear in it is optimal when, predict average every kilometer obtained etesian traffic Accident quantity, i.e., basic accident rate (secondary/km)；
βj--- linear index correction factor.
Wherein, basic accident prediction model (i.e. NB) is only relevant with annual average daily traffic (AADT).
From statistical result, when highway route is in preferable alignment condition, accident rate is with the volume of traffic (AADT) Increase quickly increase.Statistics obtains accident rate and the functional relation of annual average daily traffic, i.e., basic accident prediction model, sees Formula 2 is to formula 4.
Mountainous expressway：
NB=1.0 × 10-6×AADT1.411(R2=0.65) (2)
Hills area highway：
NB=2.06 × 10-7×AADT1.520(R2=0.69) (3)
Expressway in Plain：
NB=2.2 × 10-9×AADT2-3.75×10-5×AADT+0.96(R2=0.85) (4)
(2) correction factor of basic accident prediction model
By the preferable alignment condition defined, length of straigh line correction factor β is determinedLZ, radius of horizontal curve correction factor βRP、 Horizontal curve drift angle correction factor βα, longitudinal slope gradient correction factor βiAnd radius of vertical curve correction factor βRS。
The correction factor of each linear index.
βRP=NiRP/NiBRP (5)
βLT=NiLT/NiBLT (6)
βα=Niα/(NiBα×βRP) (7)
βi=Nii/NiBi×βRP×βα×βLT (8)
βRS=NiRS/(NiBi×βRP×βα×βLT) (9)
(3) expressway traffic accident forecast model collects
The product of accident rate forecast model and each linear index correction factor based on accident prediction model.Ground with reference to foregoing Study carefully achievement, mountainous terrain, hills area, the Expressway in Plain accident prediction model of foundation are shown in formula 10 to formula 26.
Mountainous expressway accident prediction model：
N=NB×βRP×βLZ×βα×βi×βRS (10)
NB=1.0 × 10-6×AADT1.411(R2=0.65) (11)
βRP=1248.1RP-1.041+1(R2=0.62) (12)
βLZ=0.264LZ2-0.587LZ+1.44(R2=0.65) (13)
βα=-9.25 × 10-6×α3+9.38×10-4×α2-2.38×10-2×α+1.204(R2=0.70) (14)
βi=0.072i2-0.081i+1.226(R2=0.71) (15)
βRS=12.54RS-0.46+1(R2=0.57) (16)
Hills area expressway traffic accident forecast model：
N=NB×βRP×βLZ×βα×βi×βRS (17)
NB=2.06 × 10-7×AADT1.520(R2=0.66) (18)
βRP=18755RP-1.51+1(R2=0.73) (19)
βLZ=0.373LZ2-0.758LZ+1.49(R2=0.62) (20)
βα=-1.19 × 10-5×α3+1.89×10-3×α2-0.08α+1.93(R2=0.72) (21)
βi=0.138i2+0.019i+0.99(R2=0.68) (22)
βRS=11.77RS-0.43+1(R2=0.71) (23)
Expressway in Plain accident prediction model：
N=NB×βLZ (24)
NB=2.2 × 10-9×AADT2-3.75×10-5×AADT+0.96(R2=0.85) (25)
βLZ=0.247LZ2-0.883LZ+1.91(R2=0.60) (26)
Step 2: choose linear index；
Sensitivity analysis, be exactly that hypothesized model is expressed as y=f (x1, x2 ..., xn), wherein xi be i-th of model from Variable, make each independent variable be changed in possible span, study and predict that the variation of these independents variable exports to model The influence degree of value, and the size of influence degree is referred to as to the sensitivity coefficient of the independent variable.Sensitivity coefficient is bigger, independent variable Influence to model output is bigger.In brief, sensitivity analysis is exactly that a kind of quantitative description mode input variable becomes to output A kind of method of the significance level of amount.Index selects, and exactly selects the independent variable to model, that is, have an impact to model output Variable.
For expressway traffic accident forecast model, its linear index is numerous, such as length of straigh line (LZ), horizontal curve half Footpath (RP), horizontal curve drift angle (α), the longitudinal slope gradient (i) and radius of vertical curve (RS) etc..
Step 3: calculate the sensitivity coefficient △ y of each linear indexi；
According to the scope of sensitivity analysis, part and global sensitivity analysis can be classified as.Local sensitivity analysis Only examine influence degree of the single independent variable to model, and global sensitivity analysis, examine multiple independents variable to produce model result Raw total influence, and analyze the influence that the interaction between independent variable exports to model.Local sensitivity analysis is being counted because of it Simple and fast in terms of calculation, there is very strong operability.
According to the principle of local sensitivity analysis, sensitivity coefficient calculating is carried out to the linear index selected by step 1. Some linear index is increased or is reduced in possible span, its change can be obtained accident prediction model is predicted The change of value, and then obtain the sensitivity coefficient △ y of each linear indexi, f (x1,x2,…,xn) it is that accident rate is pre- in step 1 Model is surveyed, sees formula 27.
F (x in formulai) --- accident prediction model, xiFor the different values of accident prediction model independent variable；
M --- the span of independent variable is divided into m equal portions；
△yi--- sensitivity coefficient.
Step 4: determine crucial linear index.
Crucial linear index refers to there is the prominent linear index influenceed on expressway traffic accident rate, i.e. sensitivity coefficient is maximum Linear index.From expressway traffic accident forecast model, the correction factor of linear index is bigger, expressway traffic accident rate Predicted value is also bigger.When highway route index takes different value, based on correction factor minimum value, accident rate change can be obtained Rate.Can be identified using the method for local sensitivity analysis has prominent influence to accident rate in expressway traffic accident forecast model Crucial linear index.
Step 5: considering traffic safety, the growth factor of the accident rate when the value of linear index changes is calculated.
It is worth based on the correction factor minimum value tried to achieve, and then is tried to achieve when the value of linear index changes current events Therefore the growth factor of rate, its calculation formula are shown in formula 28.
ξ in formula --- accident rate growth factor；
β --- linear index correction factor.
Step 6: determining the span recommended value of the crucial linear index in section, the crucial linear index after optimization is chosen.
The average of accident rate growth factor in calculation procedure four, and using the average of accident rate growth factor as acceptable peace The complete horizontal upper limit, on this basis, it is determined that the span of the flat vertical crucial linear index in combination section.
Beneficial effects of the present invention are verified using following examples：
Embodiment one：
A kind of new crucial linear index identification of the present embodiment is specifically to be prepared according to following steps with optimization method：
This example is based on Freeway in Liaoning Province data, according to object of this investigation and the geometrical line in binding tests section Shape condition, carry out crucial linear index selection and optimization.When highway route index takes different value, with correction factor minimum value Based on, accident rate rate of change can be obtained and be shown in Table -1.It is pre- that expressway traffic accident can be identified using the method for local sensitivity analysis Survey has the prominent crucial linear index influenceed to accident rate in model, the sensitivity coefficient of these crucial linear indexs is shown in Table -2.
From table -2：
(1) for Mountainous expressway, the sensitivity coefficient of the longitudinal slope gradient is maximum, it can thus be assumed that the longitudinal slope gradient is Influence the crucial linear index of Mountainous expressway accident rate；
(2) for the highway of hills area, the sensitivity coefficient of length of straigh line is maximum, it is believed that length of straigh line is Influence the crucial linear index of hills area expressway traffic accident rate；
(3) for Expressway in Plain, because length of straigh line is Expressway in Plain accident prediction model Unique linear target variable, thus, it is believed that length of straigh line is the crucial linear finger for influenceing Expressway in Plain accident rate Mark.
The highway route index correction factor of table 1 and accident rate rate of change
The sensitivity coefficient of each linear index of the highway of table 2
Using the vertical combination section of Curve of Freeway as research object, the crucial linear finger in the vertical combination section of Curve of Freeway have studied When marking (radius of horizontal curve, the longitudinal slope gradient, radius of vertical curve etc.) different values, the situation of change of accident rate, finally draw and be based on The recommendation of the crucial linear index span in the vertical combination section of the Curve of Freeway of traffic safety.
The accident rate growth factor in the vertical combination section of Mountainous expressway and hills area Curve of Freeway be shown in Table respectively 3 to Table 6.
From table 3 to table 6：
(1) growth factor of accident rate increases with the reduction of radius of horizontal curve, increases with the reduction of radius of vertical curve, Increase with the increase of the longitudinal slope gradient；
(2) for Mountainous expressway and hills area highway, when radius of horizontal curve takes 250m (desin speeds Limiting value during 80km/h) or during 400m (limiting value during desin speed 100km/h), no matter longitudinal slope gradient value size, perpendicular The growth factor of sweep value size accident rate is in high value；
(3), can be by increasing radius of horizontal curve when the value of radius of vertical curve is limited by orographic condition takes limiting value Size the growth factor of accident rate is reduced to lower value (average)；When radius of horizontal curve takes limiting value, increasing can not be passed through Add radius of vertical curve that the growth factor of accident rate is reduced into lower value.
The Mountainous expressway horizontal curve of table 3 combines the growth factor of section accident rate with longitudinal slope
Note：For the growth factor of accident rate combination section is indulged less than the flat of average.
The hills area highway horizontal curve of table 4 combines the growth factor of section accident rate with longitudinal slope
Note：For the growth factor of accident rate combination section is indulged less than the flat of average.The gradient is gone up a slope for negative indication, the gradient Descending is represented to be positive.
The Mountainous expressway horizontal curve of table 5 combines the growth factor of section accident rate with vertical curve
Note：For the growth factor of accident rate combination section is indulged less than the flat of average.
The hills area highway horizontal curve of table 6 combines the growth factor of section accident rate with vertical curve
Note：For the growth factor of accident rate combination section is indulged less than the flat of average.
If using the average of accident rate growth factor as the upper limit of acceptable level of security, on this basis, it is determined that The span recommended value of the flat vertical crucial linear index in combination section is as follows：
(1) horizontal curve combines the linear index span in section with longitudinal slope
When Mountainous expressway longitudinal slope is 3%, radius of horizontal curve value suggestion should be greater than 1500m；When longitudinal slope is 4%, Radius of horizontal curve value suggestion should be greater than 2500m；When longitudinal slope is 5%, radius of horizontal curve value suggestion should be greater than 6500m.
When hills area highway longitudinal slope is 3%, radius of horizontal curve value suggestion should be greater than 1000m；When longitudinal slope is 4%, Radius of horizontal curve value suggestion should be greater than 1500m；When longitudinal slope is 5%, radius of horizontal curve value suggestion should be greater than 7500m.
(2) horizontal curve combines the linear index span in section with vertical curve
When Mountainous expressway radius of vertical curve is less than 6000m, radius of horizontal curve value suggestion should be greater than 2000m.Mound When mound area highway radius of vertical curve is less than 17000m, radius of horizontal curve value suggestion should be greater than 1500m.
The span recommended value of the flat vertical crucial linear index in combination section, is shown in Table 7.
The span recommended value of the flat vertical crucial linear index in combination section of table 7
The present invention can also have other various embodiments, in the case of without departing substantially from spirit of the invention and its essence, this area Technical staff works as can make various corresponding changes and deformation according to the present invention, but these corresponding changes and deformation should all belong to The protection domain of appended claims of the invention.

Claims (5)

  1. A kind of 1. linear index selection and optimization method based on sensitivity analysis and traffic safety, it is characterised in that including：
    Step 1: according to the lineament of targeted cache highway, targeted cache highway is divided into k section；The type bag in section Include Mountainous expressway, hills area highway and Expressway in Plain；
    Step 2: for each section, the average every kilometer etesian traffic accident number that the link prediction obtains is calculated Amount；
    Step 3: set linear index correction factor for each section；Linear index correction factor is repaiied including length of straigh line Positive coefficient βLZ, radius of horizontal curve correction factor βRP, horizontal curve drift angle correction factor βα, longitudinal slope gradient correction factor βiAnd perpendicular song Line radius correction coefficient βRS；
    Step 4: according to the type in each section, the average every kilometer etesian traffic accident that the link prediction is obtained Quantity is multiplied by all linear index correction factors in the section, obtains the accident prediction model in the section；
    Step 5: length of straigh line, radius of horizontal curve, horizontal curve drift angle, the longitudinal slope gradient and radius of vertical curve are chosen as linear Index, and calculate according to the type in section the sensitivity coefficient of each linear index respectively, the sensitivity coefficient are used to describing linear Influence degree of the index to accident rate predicted value；
    Step 6: choosing the minimum linear index of sensitivity coefficient, the growth factor of the accident rate when linear index changes is sought；
    Step 7: the high linear indication range of degree of safety is filtered out according to the relation of linear index and accident rate growth factor.
  2. 2. the linear index selection and optimization method according to claim 1 based on sensitivity analysis and traffic safety, its It is characterised by, in step 2,
    When section is mountain ridge highway, is calculated by equation below and predict average every kilometer obtained etesian traffic Accident quantity NB：
    NB=1.0 × 10-6×AADT1.411
    Wherein AADT represents annual average daily traffic；
    When section is hills highway, is calculated by equation below and predict average every kilometer obtained etesian traffic Accident quantity NB：
    NB=2.06 × 10-7×AADT1.520
    When section is Expressway in Plain, is calculated by equation below and predict average every kilometer obtained etesian friendship Interpreter's event quantity NB：
    NB=2.2 × 10-9×AADT2-3.75×10-5×AADT+0.96。
  3. 3. the linear index selection and optimization method according to claim 2 based on sensitivity analysis and traffic safety, its It is characterised by, in step 4,
    When section is Mountainous expressway, the accident prediction model in the section is：
    N=NB×βRP×βLZ×βα×βi×βRS
    NB=1.0 × 10-6×AADT1.411
    βRP=1248.1RP-1.041+1
    βLZ=0.264LZ2-0.587LZ+1.44
    βα=-9.25 × 10-6×α3+9.38×10-4×α2-2.38×10-2×α+1.204
    βi=0.072i2-0.081i+1.226
    βRS=12.54RS-0.46+1
    When section is hills area highway, the accident prediction model in the section is：
    N=NB×βRP×βLZ×βα×βi×βRS
    NB=2.06 × 10-7×AADT1.520
    βRP=18755RP-1.51+1
    βLZ=0.373LZ2-0.758LZ+1.49
    βα=-1.19 × 10-5×α3+1.89×10-3×α2-0.08α+1.93
    βi=0.138i2+0.019i+0.99
    βRS=11.77RS-0.43+1
    When section is Expressway in Plain, the accident prediction model in the section is：
    N=NB×βLZ
    NB=2.2 × 10-9×AADT2-3.75×10-5×AADT+0.96
    βLZ=0.247LZ2-0.883LZ+1.91
    Wherein, LZ is length of straigh line, RP is radius of horizontal curve, α is horizontal curve drift angle, i is the longitudinal slope gradient, RS is vertical curve half Footpath.
  4. 4. the linear index selection and optimization method according to claim 3 based on sensitivity analysis and traffic safety, its It is characterised by, in step 5, sensitivity coefficient △ yi′Expression formula be：
    <mrow> <msub> <mi>&amp;Delta;y</mi> <msup> <mi>i</mi> <mo>&amp;prime;</mo> </msup> </msub> <mo>=</mo> <mfrac> <mn>1</mn> <mi>m</mi> </mfrac> <munderover> <mo>&amp;Sigma;</mo> <mrow> <msup> <mi>i</mi> <mo>&amp;prime;</mo> </msup> <mo>=</mo> <mn>1</mn> </mrow> <mi>m</mi> </munderover> <mfrac> <mrow> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>x</mi> <msup> <mi>i</mi> <mo>&amp;prime;</mo> </msup> </msub> <mo>)</mo> </mrow> <mo>-</mo> <mi>m</mi> <mi>i</mi> <mi>n</mi> <mo>{</mo> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mn>1</mn> </msub> <mo>)</mo> </mrow> <mo>,</mo> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mn>2</mn> </msub> <mo>)</mo> </mrow> <mo>,</mo> <mo>...</mo> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mi>m</mi> </msub> <mo>)</mo> </mrow> <mo>}</mo> </mrow> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> <mo>{</mo> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mn>1</mn> </msub> <mo>)</mo> </mrow> <mo>,</mo> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mn>2</mn> </msub> <mo>)</mo> </mrow> <mo>,</mo> <mo>...</mo> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>x</mi> <mi>m</mi> </msub> <mo>)</mo> </mrow> <mo>}</mo> </mrow> </mfrac> <mo>,</mo> <msup> <mi>i</mi> <mo>&amp;prime;</mo> </msup> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mo>,</mo> <mi>m</mi> </mrow>
    Wherein m represents to have carried out section m sampling, xi′Represent the sampled value to the i-th ' secondary sampling of progress of linear index, f (xi′) Represent accident rate rate of change corresponding with the sampled value of the i-th ' secondary sampling, min { f (x1),f(x2),…f(xm) represent from all M sampling corresponding to take minimum value in accident rate of change.
  5. 5. the linear index selection and optimization method according to claim 4 based on sensitivity analysis and traffic safety, its It is characterised by, in step 6, the calculation formula of accident rate growth factor is：
    <mrow> <msub> <mi>&amp;xi;</mi> <mrow> <msup> <mi>i</mi> <mrow> <mo>&amp;prime;</mo> <mo>&amp;prime;</mo> </mrow> </msup> <mi>j</mi> </mrow> </msub> <mo>=</mo> <mfrac> <mrow> <msub> <mi>&amp;beta;</mi> <mrow> <mn>1</mn> <msup> <mi>i</mi> <mrow> <mo>&amp;prime;</mo> <mo>&amp;prime;</mo> </mrow> </msup> </mrow> </msub> <mo>&amp;CenterDot;</mo> <msub> <mi>&amp;beta;</mi> <mrow> <mn>2</mn> <mi>j</mi> </mrow> </msub> </mrow> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> <mrow> <mo>(</mo> <msub> <mi>&amp;beta;</mi> <mrow> <mn>1</mn> <msup> <mi>i</mi> <mrow> <mo>&amp;prime;</mo> <mo>&amp;prime;</mo> </mrow> </msup> </mrow> </msub> <mo>&amp;CenterDot;</mo> <msub> <mi>&amp;beta;</mi> <mrow> <mn>2</mn> <mi>j</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> </mfrac> </mrow>
    Wherein ξi″jFor accident rate growth factor；β1i″For in longitudinal slope gradient correction factor i-th " individual data point, i "=1,2,3, 4……；β2jFor j-th of data point in radius of horizontal curve correction factor, j=1,2,3,4 ....
CN201710592024.4A 2017-07-19 2017-07-19 Linear index selection and optimization method based on sensitivity analysis and driving safety Active CN107392373B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710592024.4A CN107392373B (en) 2017-07-19 2017-07-19 Linear index selection and optimization method based on sensitivity analysis and driving safety

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710592024.4A CN107392373B (en) 2017-07-19 2017-07-19 Linear index selection and optimization method based on sensitivity analysis and driving safety

Publications (2)

Publication Number Publication Date
CN107392373A true CN107392373A (en) 2017-11-24
CN107392373B CN107392373B (en) 2021-02-09

Family

ID=60336424

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710592024.4A Active CN107392373B (en) 2017-07-19 2017-07-19 Linear index selection and optimization method based on sensitivity analysis and driving safety

Country Status (1)

Country Link
CN (1) CN107392373B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110264724A (en) * 2019-07-09 2019-09-20 哈尔滨工业大学 A kind of interactive high rate road traffic accident prediction technique

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020082806A1 (en) * 1995-01-13 2002-06-27 Kaub Alan R. Traffic safety prediction model
CN101694745A (en) * 2009-06-16 2010-04-14 同济大学 Safety detection method based on freeway geometry linear comprehensive technical indexes
CN104294720A (en) * 2014-09-26 2015-01-21 哈尔滨工业大学 Method for evaluating safety of expressway design scheme
CN105608902A (en) * 2016-03-28 2016-05-25 辽宁省交通科学研究院 Expressway black spot identification system and method

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020082806A1 (en) * 1995-01-13 2002-06-27 Kaub Alan R. Traffic safety prediction model
CN101694745A (en) * 2009-06-16 2010-04-14 同济大学 Safety detection method based on freeway geometry linear comprehensive technical indexes
CN104294720A (en) * 2014-09-26 2015-01-21 哈尔滨工业大学 Method for evaluating safety of expressway design scheme
CN105608902A (en) * 2016-03-28 2016-05-25 辽宁省交通科学研究院 Expressway black spot identification system and method

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
侯芹忠: "基于IHSDM框架的高速公路交通事故预测模型", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 *
孟祥海等: "IHSDM高速公路事故预测模型", 《交通运输工程学报》 *
孟祥海等: "高速公路平纵线形与事故率的关系及其安全性评价", 《交通信息与安全》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110264724A (en) * 2019-07-09 2019-09-20 哈尔滨工业大学 A kind of interactive high rate road traffic accident prediction technique

Also Published As

Publication number Publication date
CN107392373B (en) 2021-02-09

Similar Documents

Publication Publication Date Title
CN104866654B (en) A kind of construction method of integrated urban dynamic traffic emulation platform
CN105303832B (en) Overpass road section traffic volume congestion index computational methods based on microwave vehicle detector
CN106056308A (en) Highway tunnel operation environment safety risk automatic judgment method
CN102201021B (en) Expressway aided design system
CN101246513A (en) City fast road intercommunicated overpass simulation design system and selection method
CN101246514A (en) City fast road intercommunicated overpass simulation design system and method for establishing design model
CN107194049A (en) A kind of multi objective Grade system of tunnels and underground engineering rockfall risk
CN101739822B (en) Sensor network configuring method for regional traffic state acquisition
CN106296475A (en) Tunnels and underground engineering is dashed forward discharge disaster polymorphic type combining evidences appraisal procedure
CN102494667A (en) Characterizing method of land subsidence
CN110378551A (en) A kind of vcehicular tunnel facility military service method of evaluating performance based on big data
CN107180534B (en) The express highway section average speed estimation method of support vector regression fusion
CN104252556A (en) River classification system
CN107067729A (en) A kind of urban road traffic safety state evaluating method
Tengattini et al. Physical characteristics and resistance parameters of typical urban cyclists
CN107392373A (en) A kind of linear index selection and optimization method based on sensitivity analysis and traffic safety
CN107064010B (en) Soft clay area Road surface quality evaluation method
CN104376712A (en) Missing traffic information complementing device and method
Li et al. Traffic accident analysis based on C4. 5 algorithm in WEKA
Wang et al. Influencing factors on vehicles lateral stability on tunnel section in mountainous expressway under strong wind: A case of Xi-Han highway
Dixon et al. Estimating free-flow speeds for rural multilane highways
CN106529118A (en) Two-tuple linguistic AHP based intelligent automobile human-simulated steering control performance evaluation method
Xin et al. Construction of accident rate model for tunnel group sections of expressway in mountainous areas
Kim et al. Model of Volume-Delay Formula to assess travel time savings of underground tunnel roads
Meng et al. Research on accident prediction models for freeways in mountainous and rolling areas

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