Identifying the Effective Criteria for Prioritizing Accident Black-spots and their Measure of Importance

In an efficient transportation system, safety is the most important issue and is influenced by many factors. In a country similar to Iran, attention is mainly concentrated on engineering activities and with some physical adjustments, accident rates will reduce. Until recently, accident black-spots were identified and remedied by the experts’ judgments and a handful of statistics without taking into accounts other important factors such as geometric and traffic conditions of the road network. This paper aims to define and identify the criteria for accident black-spots, then giving a value to each criterion in order to develop a model to prioritize accident black-spots.In this manner, experts are required to choose any criteria which they think is or are important and correlate to this study. Here are 15 criteria to choose from:1. Geometric Conditions,2. Traffic Conditions,3. Physical Conditions,4. Topography,5. Weather Conditions,6. Distance from Population Centers,7. Specific Places, like tunnels, bridges, intersections and foggy areas,8. Type of Road (freeway, highway, arterial or access roads),9. Time Period (day or night),10. Accidents Costs,11. Maintenance Costs,12. Enforcements,13. Ratio of public transportation vehicles to all vehicles,14. State and Management Policies and15. Provision of radio communication and intelligent information systems.Then, experts are asked to choose some criteria for prioritization model. Having surveyed the expert's opinions, the most important criteria are achieved with regard to the responses of experts. After this process, regarding to number of responses to each criterion, importance of criteria was identified. In this way, geometric conditions, physical conditions, specific places, traffic conditions and distance from population centers, respectively had large number of responses.According to some expert's suggestions and some literature reviews, some of these criteria are more common than the others and they need to be considered in more detail such as geometric, traffic and physical characteristics. These main criteria, each have sub-criteria on which, the experts were asked to give their opinion about them. So, both main final criteria and sub-criteria are obtained at this stage. In addition, the experts are asked to give a weight coefficient based on Delphi method to each criterion, according to their experiences. These weights must be between 1and 9.Then, final weights of each criteria and sub-criteria were identified. In some MADM problems where a decision matrix does not exist, the decision maker's judgments regarding the comparison each criterion is essential and needed. This matrix is a reciprocal matrix with positive components. Now it is time to form a dual comparative matrix for each criterion. At this stage, the "Geometric Mean Method" is used to calculate for each criterion. In this way, the geometric mean of each row of the decision matrix will be calculated and the resulted vector must be normalized. At this stage the normalized weight values for each criterion, but according to the assumptions with structure of the prioritization models, each criterion and sub-criterion must be in the same level. So, to satisfy this stipulation, weight of sub-criteria must be multiplied by their correlate criteria weight. At last the final weight matrix which consists of 14 numbers was get. To do all these stages, the "Delphi" method has been used.