Traffic Safety Evaluation Model for Highway in Cold Region Based on Fuzzy Theory

Aiming at the problem of highway traffic safety in cold region, traffic safety evaluation model is established by applying the fuzzy comprehensive evaluation theory and the highway safety level is judged. On the basis of the comprehensive analysis of the factors affecting highway traffic safety, the index system for highway traffic safety evaluation is established combining with expert evaluation method, the weight value of influence factors in the traffic incident are confirmed. The comprehensive evaluating result is obtained. The value is 1.8858 in range from 1 to 2. The result shows that the highway safety level in the certain cold region is good. The assessment model can offer the theoretical foundation for highway safety evaluation in other cold region.


INTRODUCTION
With the development of economy, Highway construction has come into fast development stage.At the same time, the high-speed flow and gather of people and the rapid growth amount of vehicle make the traffic demand increasingly strong, the transportation security problems become the inevitable result of social economic development (Ma, 2012).In recent years, the highway traffic accident rate remains high and the highway traffic accidents are particularly serious in the cold region.Scholars at home and abroad have devoted to the researches of highway traffic safety evaluation and have achieved significant results.Foreign scholars propose to use Poisson regression model (Miaou and Lum, 1993), negative binomial regression model (Hinde and Demetrio, 1998), zero-inflated probability model (Lord et al., 2005), to analyze the relationship between traffic accidents and the influence factors.Zhao et al. (2012) from Chongqing Jiaotong University establish a total vehicle travel safety rating model by using the Analytic Hierarchy Process to calculate the parameters weight.Yao et al. (2010) establish safety rating evaluation system based on the three aspects of road alignment, vehicle design parameters and road-vehicle correlation parameters and build the mountain road safety rating model combining the fuzzy mathematics method.As it is difficulty to establish the scientific comprehensive model accords with current situation, Wang and Nan (2008) propose macroscopic evaluation model based on fuzzy logic.Due to the special geography, climatic conditions and the complex road environment, road traffic accidents in cold region have distinguishing features different from other parts.Traffic accident statistics analysis showed that the accidents happened in cold region had seasonal variations.In winter, traffic accident rate was obviously improved.The highway traffic safety fuzzy evaluation model established in this study is significance for reducing the traffic accident rate, which can judge the highway safety level through finding out and analyzing highway traffic safety factors in cold region.

THE FUZZY COMPREHENSIVE EVALUATION THEORY
Applying the fuzzy comprehensive evaluation theory will arrive at a scientific evaluation conclusion through selecting the neighborhood of each factor in a system reasonably and evaluating the factors.Its fundamental theory is: where, F = The total score of the system B = The evaluation matrix of the system S T = Factor fraction: where, B = The evaluation matrix of the system A = The weight distribution set of each factor R = Evaluation matrix: where, B i = Sub-factors of evaluation matrix in the system Maximum characteristic root of matrix: Consistency index of matrix: Mean random Consistency index: The steps of applying the fuzzy comprehensive evaluation are as followed.Firstly, determine quantitative criteria.Secondly, determine judgment matrix.Thirdly, Single-level sequencing and consistency check, calculate maximum characteristic root λ max of the judgment matrix A and its corresponding single ranking weight vector W and then calculate the random consistency ratio CR, carry on the consistency check: Then establish evaluation set U =｛U 1 , U 2 , U 3 , U 4 and U 5 ｝, determine the safety level of highway in cold region, the evaluation set is U = {excellent, good, medium, low-risk, high risk}, which correspond grade are 1, 2, 3, 4 and 5 respectively.

HIGHWAY SAFETY EVALUATION SYSTEM IN COLD REGION
Applying the multi-level comprehensive evaluation method to evaluate highway safety level in cold region based on many influence factors, the safety comprehensive evaluation index system for highway in the typical cold region Heilongjiang province is established.The highway safety comprehensive evaluation index system in cold region is shown in Table 1.
• Scheme evaluation of highway safety level: • Establish the evaluation matrix: Scheme evaluation result is shown in Table 7 through marking and accessing the factors related to highway safety in cold region by experts.
Each judgment matrix of environmental factor in winter, influence of low temperature on performance of automobile, driver personal factors is shown as follows according to Table7: 1<P 1 = 1.8858<2, so the highway traffic safety level in a region of Heilongjiang Province in winter is good.

CONCLUSION
This study comprehensively analyzes the factors influencing the highway traffic safety in typical cold region of Heilongjiang Province in winter for studying the problems of highway traffic safety in cold region, selects evaluation index and establishes highway traffic safety evaluation system in cold region.Highway traffic safety fuzzy comprehensive evaluation model in cold region can be established through conducting quantitative evaluation on highway traffic safety in cold region combining with the fuzzy theory , transforming the multi-objective evaluation problems of the highway traffic safety into single objective evaluation, determining the weight value of influence factors in the traffic incident.The highway traffic safety level in the region in winter is good for the comprehensive evaluation value of scheme evaluation 1<P 1 = 1.8858<2.The highway safety level evaluation index system in cold region built in this study can more accurately reflect the traffic safety condition in a region and offer theoretical foundation for highway safety evaluation in other cold region.

ACKNOWLEDGMENT
This study was supported by the National Natural Science Foundation of China (Grant No. 51108068) and the S&T Plan Projects of Heilongjiang Provincial Education Department (Grant No. 11553025) and the Fundamental Research Funds for the Central Universities (Grant No. DL12CB03).

•
Confirm fuzzy relation matrix: Conduct Onelevel fuzzy comprehensive assessment is conducted and the fuzzy relation matrix R = (R 2 , R 3 and R 5 ) T is confirmed: So the fuzzy relation matrix R can be obtained.

• Sub-factors set of evaluation factors sets: Sub
2 -C 2 level, B 3 -C 3 level, B 5 -C 5 level and consistency check are shown inTable 2, 3, 4 and 5. • General sequence and consistency check: Based on single sequence results and general sequence weight of evaluation factors, generals equence weight.

Table 2 :
Judgment matrix of A-B

Table 6 :
Weight and rank of factors

Table 7 :
Assessment grade of influencing factors of highway traffic safety in cold region