Research On Accidents Causes Based On AHP-Apriori Algorithm

Traffic safety has become a difficult problem the worldwide common concernand explore The occurrence of traffic accidents is sudden, accidental, governed by avariety of factors, it is the comprehensive result of interaction between variousinfluencing factors. If we can find the regular contain in traffic accident data, such asRoad conditions, vehicle conditions, time, weather condition and accident types,accident form contained by accident itself attribute interact with each other, dopreventive measures in advance, control the accident risk under the manageable range,it will be able to effectively prevent the happening of the accident and reduce thedamage.The road traffic management department made detailed record data for theaccident at present, built traffic accident information system, But it still appliedtraditional probability statistics method,ignored the description of traffic accidentdata whole character description which has more decision significance, Correlationbetween the data, accident occurrence regularity and development trend forecast.The accident data collected has not been fully mining and application, the madoriginal data takes up a lot of storage space, cause the waste of resources.This paper based on the road traffic accident database of the public securityministry, Using data mining technology, According to the characteristics of themultidimensional and hierarchical causes of traffic accidents, From the driver, vehicle,road and environment set out four elements, quantitative it based on objective analysis,To study the causes of traffic accidents, build the accident model based on AHP-Apriori algorithm, firstly using AHP sorting the importance of accident influencefactors, Find out the main factors affecting the accident, based on it using thealgorithm of association rules on main factors affecting accidents which have beenselected to correlation analysis. Analysis shows that, the method can determine theaccident risk factors and the space-time characteristics of the accident under differentconditions, it can improve the operation efficiency and analysis accuracy especially inthe big traffic accident data samples.Finally we have safety evaluation and warning based on the algorithm, and putforward safety management suggestions and road safety measures to guide the traffic administrative department decision, put an end to accident risks, reduce accidents, andimprove the road safety level ultimately.