CN106651191A - Mountain area highway horizontal curve dangerous section identification method - Google Patents

Abstract

The invention provides a mountain area highway horizontal curve dangerous section identification method. Analysis is carried out through utilization of specialist knowledge and related event statistic information in the field, thereby obtaining evaluation parameters of a mountain area highway horizontal curve section; a membership function of each influence factor is determined through operation of a fuzzy specialist basic theory; through combination of the specialist knowledge in the field, evaluation rules are extracted and established; a mountain area highway horizontal curve dangerous section identification fuzzy specialist system is established, thereby carrying out dangerous degree calculation and sorting on each horizontal curve section. According to the mountain area highway horizontal curve dangerous section identification system which is based on the fuzzy specialist system and is established by the method, predictive evaluation can be carried out on a road in a design stage.

Description

A kind of mountainous area highway horizontal curve Dangerous Area recognition methods

Technical field

The invention belongs to traffic safety engineering field, and in particular to a kind of mountainous area highway horizontal curve Dangerous Area Recognition methods.

Background technology

In recent years, China's freeway network rapid development, 2014, it is public that China's highway total kilometrage breaks through 12.5 ten thousand In, but because China mountain region and knob area account for more than 2/3rds of territory total area, freeway net is increasing rapidly While long, mountainous area highway mileage is also being increased rapidly.However, because the linear index of mountainous area highway is relatively low, road The reasons such as structure particular row, frequent accidents occur.Only 2011, the Road Traffic Accidents in China number of casualties close 300,000 People, is only second to India, while the ratio that Horizontal Curve Sections accident quantity accounts for sum is up to 10.5%, is only second to drive over the speed limit and causes Accident quantity.It can be seen that, it is considered to the relatively low linear index of mountainous area highway, strengthen horizontal curve Dangerous Area Prevention Research in advance Work is very urgent.

Consider that the reason for whether the road Horizontal Curve Sections under the conditions of different road geometric factors are Dangerous Area is：Vapour Car can produce centrifugal force when Horizontal Curve Sections are travelled, and it is very big to the stability influence that automobile is travelled on circular curve, may Make the outside side slip of automobile or topple, size and the circular curve radius of centrifugal force are inversely proportional to, therefore when radius is less, garage Run over unstable, be more susceptible to traffic accident；Mountainous area highway longitudinal gradient is generally larger, during automobile climbing or descending Braking more frequently, the friction plate temperature of brake is too high and cause brake failure, so as to cause traffic accident；Highways in mountain areas is public Route in landform limit, horizontal curve drift angle choose may result in it is excessive or too small, when horizontal curve drift angle it is too small, even if radius Bigger, driver can also regard horizontal curve length as than reality shorter, and the illusion taken a sudden turn is caused to it, be unfavorable for security row Car, when horizontal curve drift angle is excessive, the transformation angle of driver's seat becomes big, causes sighting distance bad and causes the generation of accident.Cause This, it is considered to the geometrical factor of mountainous area highway Horizontal Curve Sections, and further to the domain expert related investigation and analysis are carried out, To with dangerous sequence is carried out without the mountainous area highway horizontal curve of geometrical property, looking for out most dangerous section, and adopt Corresponding safety measure is taken, such that it is able to avoid the generation of a number of traffic accident.

Chinese scholar has been done on to the dangerous research of highway Horizontal Curve Sections and has been studied in large quantities, however, with Method substantially actual casualty data is analyzed, it was therefore concluded that.But, China's incident database is not perfect enough, public Peace department to accident information also not full disclosure, add places at different levels understatement for some reason to actual accidents situation, Conceal, the casualty data for causing research institute to adopt is very inaccurate；Meanwhile, actual accidents data are obtained based on existing road, But existing road is improved, is taken time and effort.Therefore, if actual accident number can not relied in the design phase According in the case of, it is considered to designed the geometrical property factor for completing road, the danger of Horizontal Curve Sections has been ranked up, not only Can be with time saving and energy saving, it is also possible to be that the result for obtaining is more accurate.

The content of the invention

The problem to be solved in the present invention is to provide a kind of mountainous area highway horizontal curve Dangerous Area recognition methods, by this Method is established based on the mountainous area highway horizontal curve Dangerous Area identifying system of fuzzy expert system, can be to being still in design The evaluation of the being predicted property of road in stage.

In order to achieve the above object, the present invention adopts below scheme：

A kind of mountainous area highway horizontal curve Dangerous Area recognition methods, comprises the following steps：

(1) it is by highway circular curve radius, road longitudinal grade, horizontal curve corner as input variable, horizontal curve is dangerous Property as output variable, it is considered to specific design data, list the value strategy of three parameters；

(2) each input value is carried out into Fuzzy processing, determines the linguistic labelses of input value；

(3) linguistic labelses determined according to each input value, determine the span of each linguistic labels of input value；

(4) because having different unit and dimension between different input and output values, therefore by each input and output value by removing It is standardized with its maximum, is specifically shown in Table 1- tables 4；

The circular curve radius of table 1 and span

The road longitudinal grade of table 2 and span

The horizontal curve corner of table 3 and span

The horizontal curve risk factor of table 4 and span

In table 1- tables 4, R represents circular curve radius, and i represents road longitudinal grade, and a represents horizontal curve corner, and d represents horizontal curve danger It is dangerous；

(5) fuzzy set of the fuzzy membership function of each input/output variable, i.e. each input/output variable is determined：

1. the membership function of circular curve radius

It is little：

Scope：[0,0,0.08,0.30]

Properly：

Scope：[0.08,0.30,0.50,0.70]

Greatly：

Scope：[0.50,0.70,1.00,1.00]

2. the membership function of road longitudinal grade

It is little：

Scope：[0,0,0.17,0.33]

Properly：

Scope：[0.17,0.33,0.50,0.83]

Greatly：

Scope：[0.50,0.83,1.00,1.00]

3. the membership function of horizontal curve corner

It is little：

Scope：[0,0,0.12,0.30]

Properly：

Scope：[0.12,0.30,0.58,0.83]

Greatly：

Scope：[0.58,0.83,1.00,1.00]

4. the membership function of horizontal curve risk factor

It is very safe：

Scope：[0,0,0.11,0.22]

Safety：

Scope：[0.11,0.22,0.33,0.44]

It is susceptible to danger：

Scope：[0.33,0.44,0.55,0.66]

It is dangerous：

Scope：[0.55,0.66,0.77,0.88]

It is abnormally dangerous：

Scope:[0.77,0.88,1.00,1.00]

(6) using the Rule Expression mode of production, fuzzy rule base is obtained；

(7) for one group of input R, i, a, any one input value R, i or a all at least belong to it is a kind of it is little, in or big language Speech label, but also can simultaneously belong to bilingual label, according to the linguistic labelses that each input value has, in fuzzy rule base It is middle to select corresponding rule to be calculated；

(8) exact value of each input value R, i, a is obtained, then according to suitable linguistic fuzzy set carries out obfuscation, it is right In standardized input value R, calculated by membership function (1)-(3) obtain certain linguistic labels be subordinate to angle value μ_RS(R)、 μ_RM(R)、μ_RL(R), in the same manner, for input value i and a, it can be calculated respectively and be subordinate to angle value μ_iS(i)、μ_iM(i)、μ_iL(i) and μ_aS(a)、μ_aM(a)、μ_aL(a)；

(9) apply them in corresponding rule after the obfuscation output for obtaining each input value, each rule includes Two parts：IF statement part and THEN statement parts, IF statement part describes the premise of fuzzy rule, and THEN statement parts are retouched The conclusion of fuzzy rule is stated, due to there are circular curve radius R, road longitudinal grade i and tri- input values of horizontal curve drift angle a, therefore IF languages Sentence part is regular using the AND of fuzzy logic, and its computing formula is as follows：

μ_R∩i∩a(x)=min [μ_R(x),μ_i(x),μ_a(x)] (15)

Min is represented and is taken μ_R(x)、μ_i(x)、μ_aThe minimum of a value of (x) three；

A numerical value for representing IF statement component assesses result is obtained after calculating, next this numerical value applied To THEN statement parts；

(10) because each input value has no linguistic labelses, corresponding rule is transferred in fuzzy rule base, by An output result is produced in a rule, therefore needs that the single result that strictly all rules is exported is merged in a fuzzy set, Form single fuzzy set；

(11) according to the single fuzzy set for being formed, the barycenter of the fuzzy set is calculated using barycenter defuzzification method, is calculated Shown in formula such as formula (16), so as to obtain clearly output valve, the value is mountainous area highway horizontal curve Dangerous Area Risk factor, so that it is determined that the degree of danger of horizontal curve

The mountainous area highway horizontal curve Dangerous Area recognition methods of the present invention, using expert of the art's knowledge and related thing Therefore statistics, analyze the evaluating for obtaining mountainous area highway Horizontal Curve Sections；Determine with fuzzy expert basic theories The membership function of each influence factor, then, with reference to expertise in the art, extracts and sets up evaluation rule, and finally Setting up mountainous area highway horizontal curve Dangerous Area identification fuzzy expert system carries out dangerous big subtotal to each Horizontal Curve Sections Calculate and sequence, the mountainous area highway horizontal curve Dangerous Area identification system based on fuzzy expert system is established by this method System, can be to being still in the evaluation of the being predicted property of road of design phase.

Description of the drawings

Fig. 1 is the membership function image of input and output value

The degree of membership image of (a) circular curve radius

The degree of membership image of (b) road longitudinal grade

The degree of membership image of (c) horizontal curve corner

The degree of membership image of (d) horizontal curve risk factor

Fig. 2 is based on the Dangerous Area reasoning process of fuzzy expert system；

Fig. 3 is the flow chart of the present invention；

Fig. 4 Different Rule composition of fuzzy relations figures；

Fig. 5 is the fuzzy set after synthesis

Specific embodiment

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete Site preparation is described, it is clear that described embodiment a part of embodiment only of the invention, rather than whole embodiments, is based on Embodiment in the present invention, those of ordinary skill in the art do not making obtained under the premise of creative work it is all its His embodiment, belongs to the scope of protection of the invention.

To become apparent from the object, technical solutions and advantages of the present invention, below in conjunction with the accompanying drawings to the concrete reality of the present invention Example elaborates.

With reference to Fig. 1-3, a kind of mountainous area highway horizontal curve Dangerous Area based on fuzzy expert system of the present invention is recognized Method, as shown in scheme 3, comprises the steps of：

(1)：Western 3 Horizontal Curve Sections of certain highways in mountain areas are chosen as research object, its circular curve radius, road is extracted Longitudinal gradient, horizontal curve corner, see the table below：

Numbering	Circular curve radius R/m	I/ ° of road longitudinal grade	Horizontal curve corner α/°
				1	1035	4.5	27
2	3255	1.26	39.18
				3	2100	2.7	33

(2)-(4)：According to table 1- tables 4, it is determined that the linguistic labelses and standardized value of 3 Horizontal Curve Sections chosen.

By taking section 1 as an example：

The circular curve radius of table 1 and span

The road longitudinal grade of table 2 and span

The horizontal curve corner of table 3 and span

(5)：Linguistic labelses and scope according to belonging to circular curve radius, road longitudinal grade peace curvilinear corner value, substitute into phase The membership function answered is calculated.By taking section 1 as an example, radius of horizontal curve should be calculated using formula (1) (2), and road is indulged Slope should be calculated using formula (5) (6), and horizontal curve corner is calculated using formula (8).

(6)：By investigation and assayer's knowledge, using the Rule Expression mode of production, following fuzzy rule is obtained Storehouse：

1.IF(Radius of horizontal curve is S)and(Road Slope is S)and (Declination Angle is S)THEN(the danger level of road section is D)

2.IF(Radius of horizontal curve is S)and(Road Slope is M)and (Declination Angle is S)THEN(the danger level of road section is D)

3.IF(Radius of horizontal curve is S)and(Road Slope is L)and (Declination Angle is S)THEN(the danger level of road section is VD)

4.IF(Radius of horizontal curve is S)and(Road Slope is S)and (Declination Angle is M)THEN(the danger level of road section is S)

5.IF(Radius of horizontal curve is S)and(Road Slope is S)and (Declination Angle is M)THEN(the danger level of road section is P)

6.IF(Radius of horizontal curve is S)and(Road Slope is L)and (Declination Angle is M)THEN(the danger level of road section is D)

7.IF(Radius of horizontal curve is S)and(Road Slope is S)and (Declination Angle is L)THEN(the danger level of road section is D)

8.IF(Radius of horizontal curve is S)and(Road Slope is M)and (Declination Angle is L)THEN(the danger level of road section is D)

9.IF(Radius of horizontal curve is S)and(Road Slope is L)and (Declination Angle is L)THEN(the danger level of road section is VD)

10.IF(Radius of horizontal curve is M)and(Road Slope is S)and (Declination Angle is S)THEN(the danger level of road section is P)

11.IF(Radius of horizontal curve is M)and(Road Slope is M)and (Declination Angle is S)THEN(the danger level of road section is P)

12.IF(Radius of horizontal curve is M)and(Road Slope is L)and (Declination Angle is S)THEN(the danger level of road section is D)

13.IF(Radius of horizontal curve is M)and(Road Slope is S)and (Declination Angle is M)THEN(the danger level of road section is VS)

14.IF(Radius of horizontal curve is M)and(Road Slope is M)and (Declination Angle is M)THEN(the danger level of road section is S)

15.IF(Radius of horizontal curve is M)and(Road Slope is L)and (Declination Angle is M)THEN(the danger level of road section is P)

16.IF(Radius of horizontal curve is M)and(Road Slope is S)and (Declination Angle is L)THEN(the danger level of road section is P)

17.IF(Radius of horizontal curve is M)and(Road Slope is M)and (Declination Angle is L)THEN(the danger level of road section is P)

18.IF(Radius of horizontal curve is M)and(Road Slope is L)and (Declination Angle is L)THEN(the danger level of road section is D)

19.IF(Radius of horizontal curve is L)and(Road Slope is S)and (Declination Angle is S)THEN(the danger level of road section is P)

20.IF(Radius of horizontal curve is L)and(Road Slope is M)and (Declination Angle is S)THEN(the danger level of road section is S)

21.IF(Radius of horizontal curve is L)and(Road Slope is L)and (Declination Angle is S)THEN(the danger level of road section is P)

22.IF(Radius of horizontal curve is L)and(Road Slope is S)and (Declination Angle is M)THEN(the danger level of road section is VS)

23.IF(Radius of horizontal curve is L)and(Road Slope is M)and (Declination Angle is M)THEN(the danger level of road section is VS)

24.IF(Radius of horizontal curve is L)and(Road Slope is L)and (Declination Angle is M)THEN(the danger level of road section is S)

25.IF(Radius of horizontal curve is L)and(Road Slope is S)and (Declination Angle is L)THEN(the danger level of road section is S)

26.IF(Radius of horizontal curve is L)and(Road Slope is M)and (Declination Angle is L)THEN(the danger level of road section is S)

27.IF(Radius of horizontal curve is L)and(Road Slope is L)and (Declination Angle is L)THEN(the danger level of road section is D)

(7)：According to 27 of fuzzy rule base, therefrom choose suitable rule and matched.By taking section 1 as an example, song is put down The linguistic labelses of line radius are little and suitable, and the linguistic labelses of road longitudinal grade are suitable and big, the linguistic labelses of horizontal curve corner For suitable, therefore selection rule 5, rule 6, rule 14, regular 15 4 rules carry out the meter of next step from fuzzy rule base Calculate.

(8)：According to the linguistic labelses that input value has, formula (1)-(9) are selected to carry out the calculating of degree of membership.With section As a example by 1, calculated using formula (1), (2), (5), (6), (8), obtain μ_RS(R=0.207)=0.423, μ_RM(R=0.207)= 0.577、μ_iM(i=0.75)=0.242, μ_iL(i=0.75)=0.756, μ_aM(a=0.45)=1.

(9)：Rule according to selecting carries out the calculating of output valve d.By taking section 1, rule 5 as an example：

μ_RS∩iM∩aM(x)=min [μ_RS(R=0.207), μ_iM(i=0.75), μ_aM(a=0.45)]

=min [0.423,0.756,1]

=0.423

I.e.：If it is that the degree of danger in section is in the case of little, horizontal curve corner is suitable that radius is little, longitudinal gradient：Hold Easily cause danger, its degree for being susceptible to danger is 42.3%.

(10)：According to the rule that input value is matched, the output of each rule is merged into a not fuzzy set, formed single Fuzzy set.By taking section 1 as an example, the single fuzzy set of output is as shown in Figure 4：

(11)：The barycenter of fuzzy set after synthesis is calculated according to formula (16), result of calculation is the danger of the Horizontal Curve Sections Dangerous degree.Fig. 5 is the fuzzy set after synthesis.

Therefore, it is possible to judge that the risk factor of Horizontal Curve Sections 1 is 52.5%, in the same manner section 2 can be calculated with the method Risk factor is 24.2%, and the risk factor in section 3 is 27.5%.

Claims (1)

1. a kind of mountainous area highway horizontal curve Dangerous Area recognition methods, it is characterised in that comprise the following steps：

(1) highway circular curve radius, road longitudinal grade, horizontal curve corner are worked as horizontal curve danger as input variable Make output variable, it is considered to specific design data, list the value strategy of three parameters；

(2) each input value is carried out into Fuzzy processing, determines the linguistic labelses of input value；

(3) linguistic labelses determined according to each input value, determine the span of each linguistic labels of input value；

(4) because having different unit and dimension between different input and output values, therefore by each input and output value by divided by it Maximum is standardized, and is specifically shown in Table 1- tables 4；

The circular curve radius of table 1 and span

The road longitudinal grade of table 2 and span

The horizontal curve corner of table 3 and span

The horizontal curve risk factor of table 4 and span

In table 1- tables 4, R represents circular curve radius, and i represents road longitudinal grade, and a represents horizontal curve corner, and it is dangerous that d represents horizontal curve Property；

(5) fuzzy set of the fuzzy membership function of each input/output variable, i.e. each input/output variable is determined：

1. the membership function of circular curve radius

It is little：

Scope：[0,0,0.08,0.30]

${\mathrm{\&mu;}}_{RS}\left(x\right)=\left\{\begin{array}{cc}1& 0<x\&le;0.08\\ \frac{0.3-x}{0.22}& 0.08<x\&le;0.3\\ 0& 0.3<x\&le;1\end{array}\right.---\left(1\right)$

Properly：

Scope：[0.08,0.30,0.50,0.70]

${\mathrm{\&mu;}}_{RM}\left(x\right)=\left\{\begin{array}{cc}0& 0<x\&le;0.08\\ \frac{x-0.08}{0.22}& 0.08<x\&le;0.3\\ 1& 0.3<x\&le;0.5\\ \frac{0.7-x}{0.2}& 0.5<x\&le;0.7\\ 0& 0.7<x\&le;1\end{array}\right.---\left(2\right)$

Greatly：

Scope：[0.50,0.70,1.00,1.00]

${\mathrm{\&mu;}}_{RL}\left(x\right)=\left\{\begin{array}{cc}0& 0<x\&le;0.5\\ \frac{x-0.5}{0.2}& 0.5<x\&le;0.7\\ 1& 0.7<x\&le;1\end{array}\right.---\left(3\right)$

2. the membership function of road longitudinal grade

It is little：

Scope：[0,0,0.17,0.33]

${\mathrm{\&mu;}}_{iS}\left(x\right)=\left\{\begin{array}{cc}1& 0<x\&le;0.17\\ \frac{0.83-x}{0.66}& 0.17<x\&le;0.83\\ 0& 0.83<x\&le;1\end{array}\right.---\left(4\right)$

Properly：

Scope：[0.17,0.33,0.50,0.83]

${\mathrm{\&mu;}}_{iM}\left(x\right)=\left\{\begin{array}{cc}0& 0<x\&le;0.17\\ \frac{x-0.17}{0.16}& 0.17<x\&le;0.33\\ 1& 0.33<x\&le;0.5\\ \frac{0.83-x}{0.33}& 0.5<x\&le;0.83\\ 0& 0.83<x\&le;1\end{array}\right.---\left(5\right)$

Greatly：

Scope：[0.50,0.83,1.00,1.00]

${\mathrm{\&mu;}}_{iL}\left(x\right)=\left\{\begin{array}{cc}0& 0<x\&le;0.5\\ \frac{x-0.5}{0.33}& 0.5<x\&le;0.83\\ 1& 0.83<x\&le;1\end{array}\right.---\left(6\right)$

3. the membership function of horizontal curve corner

It is little：

Scope：[0,0,0.12,0.30]

${\mathrm{\&mu;}}_{aS}\left(x\right)=\left\{\begin{array}{cc}1& 0<x\&le;0.12\\ \frac{0.3-x}{0.18}& 0.12<x\&le;0.3\\ 0& 0.3<x\&le;1\end{array}\right.---\left(7\right)$

Properly：

Scope：[0.12,0.30,0.58,0.83]

${\mathrm{\&mu;}}_{aM}\left(x\right)=\left\{\begin{array}{cc}0& 0<x\&le;0.12\\ \frac{x-0.12}{0.18}& 0.12<x\&le;0.3\\ 1& 0.3<x\&le;0.58\\ \frac{0.83-x}{0.25}& 0.58<x\&le;0.83\\ 0& 0.83<x\&le;1\end{array}\right.---\left(8\right)$

Greatly：

Scope：[0.58,0.83,1.00,1.00]

${\mathrm{\&mu;}}_{aL}\left(x\right)=\left\{\begin{array}{cc}0& 0<x\&le;0.58\\ \frac{x-0.58}{0.25}& 0.58<x\&le;0.83\\ 1& 0.83<x\&le;1\end{array}\right.---\left(9\right)$

4. the membership function of horizontal curve risk factor

It is very safe：

Scope：[0,0,0.11,0.22]

${\mathrm{\&mu;}}_{dVS}\left(x\right)=\left\{\begin{array}{cc}1& 0<x\&le;0.11\\ \frac{0.22-x}{0.11}& 0.11<x\&le;0.22\\ 0& 0.22<x\&le;1\end{array}\right.---\left(10\right)$

Safety：

Scope：[0.11,0.22,0.33,0.44]

${\mathrm{\&mu;}}_{dS}\left(x\right)=\left\{\begin{array}{cc}0& 0<x\&le;0.11\\ \frac{x-0.11}{0.11}& 0.11<x\&le;0.22\\ 1& 0.22<x\&le;0.33\\ \frac{0.44-x}{0.11}& 0.33<x\&le;0.44\\ 0& 0.44<x\&le;1\end{array}\right.---\left(11\right)$

It is susceptible to danger：

Scope：[0.33,0.44,0.55,0.66]

${\mathrm{\&mu;}}_{dP}\left(x\right)=\left\{\begin{array}{cc}0& 0<x\&le;0.33\\ \frac{x-0.33}{0.11}& 0.33<x\&le;0.44\\ 1& 0.44<x\&le;0.55\\ \frac{0.66-x}{0.11}& 0.55<x\&le;0.66\\ 0& 0.66<x\&le;1\end{array}\right.---\left(12\right)$

It is dangerous：

Scope：[0.55,0.66,0.77,0.88]

${\mathrm{\&mu;}}_{dD}\left(x\right)=\left\{\begin{array}{cc}0& 0<x\&le;0.55\\ \frac{x-0.55}{0.11}& 0.55<x\&le;0.66\\ 1& 0.66<x\&le;0.77\\ \frac{0.88-x}{0.11}& 0.77<x\&le;0.88\\ 0& 0.88<x\&le;1\end{array}\right.---\left(13\right)$

It is abnormally dangerous：

Scope:[0.77,0.88,1.00,1.00]

${\mathrm{\&mu;}}_{dVD}\left(x\right)=\left\{\begin{array}{cc}0& 0<x\&le;0.77\\ \frac{x-0.77}{0.11}& 0.77<x\&le;0.88\\ 1& 0.88<x\&le;1\end{array}\right.---\left(14\right)$

(6) using the Rule Expression mode of production, fuzzy rule base is obtained；

(7) for one group of input R, i, a, any one input value R, i or a all at least belong to it is a kind of it is little, in or big language mark Sign, but also can simultaneously belong to bilingual label, according to the linguistic labelses that each input value has, select in fuzzy rule base Select corresponding rule to be calculated；

(8) exact value of each input value R, i, a is obtained, then according to suitable linguistic fuzzy set carries out obfuscation, for mark Input value R of standardization, is calculated by membership function (1)-(3) and obtains certain linguistic labels and be subordinate to angle value μ_RS(R)、μ_RM (R)、μ_RL(R), in the same manner, for input value i and a, it can be calculated respectively and be subordinate to angle value μ_iS(i)、μ_iM(i)、μ_iL(i) and μ_aS (a)、μ_aM(a)、μ_aL(a)；

(9) apply them to after the obfuscation output for obtaining each input value in corresponding rule, each rule includes two Part：IF statement part and THEN statement parts, IF statement part describes the premise of fuzzy rule, and THEN statement parts describe mould The conclusion of paste rule, due to there are circular curve radius R, road longitudinal grade i and tri- input values of horizontal curve drift angle a, therefore IF statement portion Point using fuzzy logic AND it is regular, its computing formula is as follows：

μ_R∩i∩a(x)=min [μ_R(x),μ_i(x),μ_a(x)] (15)

Min is represented and is taken μ_R(x)、μ_i(x)、μ_aThe minimum of a value of (x) three；

A numerical value for representing IF statement component assesses result is obtained after calculating, next this numerical value be applied to THEN statement parts；

(10) because each input value has no linguistic labelses, corresponding rule is transferred in fuzzy rule base, due to one Rule produces an output result, therefore needs that the single result that strictly all rules is exported is merged in a fuzzy set, is formed Single fuzzy set；

(11) according to the single fuzzy set for being formed, the barycenter of the fuzzy set, computing formula are calculated using barycenter defuzzification method As shown in formula (16), so as to obtain clearly output valve, the value is the danger of mountainous area highway horizontal curve Dangerous Area Degree, so that it is determined that the degree of danger of horizontal curve

$d=\frac{\underset{a}{\overset{b}{\&Integral;}}{\mathrm{\&mu;}}_d\left(x\right)xdx}{\underset{a}{\overset{b}{\&Integral;}}{\mathrm{\&mu;}}_d\left(x\right)dx}---\left(16\right).$