The Research Of The More Road Traffic Accident Area Intelligent Checking System

The safety of Road traffic is a social issue in the world. The death toll has exceeded that of war caused by traffic accident, economic loss is about 1% to 3% of total of every country's GDP.The condition of our country's traffic safety is more and more serious. Our country traffic accident has been continues increasing since the middle term of 1980'. Traffic accident death toll exceed 100,000 people over the past 2001 in a row for 3 years , the number of wounded people every year amounts to 500,000, direct economic losses already is up to more than $3,000,000,000. Our country automobile is preserved and owned measuring appointment accounting for 3% of the world, while road accident death rate accounts for 15% of the world, standing high at the world for many years. The traffic accident brings state property and the people life property safety enormous loss.The country all needs to cost large amount of manpower and financial resources every year to do lots of things such as checking the road area of more accident and establish renovate countermeasures in order to guarantee state property and the people life property to rectify and administer a countermeasure. The road accident is "about 11-5" priority the problem solving, the important road traffic ten thousand accident death rates was established the import item of checking all of the various level government and the enterprise leader during "about 11-5" formally. Country's target is that the vehicle accident death rate comes down from 7.60 in 2005 to 4.54 to 2010 with ten thousand. Therefore, pertinent intellectualized system would be researched and developed helping resolve traffic safety problem.Traffic accident process takes on system complexities and matter high characteristics for the development of modern science technology and scale economy, the improvement of mechanization, information and centralizing level, appearing of high-speed, high need and high frequencies,accident dashing forward because of nature , conceal, outburst feature and regularity become the main studying content of safe problem.The paper mainly researches how to use data excavating technologies to check the area of more accident in traffic accident database. The traffic black spot checking method based on clustering analysis algorithm is put forward after analyzing the method being formed of road accident black spot. The clustering analysis method based on density can find out biggest area of density of space. The traffic black spot is not bad look upon as is concentrated area of road accident, therefore being able to be checked by clustering analysis method.DBSCAN is a representative cluster-algorithm. Owe an algorithm the concept making use of density-based clustering , i.e., requiring that certain area inner in clustering space is that number contained a marriage partner by epsilon neighborhood is not smaller than some one given threshold value. Algorithmic merit of DBSCAN is that clustering speed is quick and it can have an effect to handle noise burning and find that the form space gathers a kind arbitrarily.The following gives out three kinds of checking pattern of more road accidents area:1. Checking pattern based on the traditional DBSCSAN algorithm; 2. Having appended examination process, the accident grand of severity has also been considered; 3. Ascertaining a initial class, and then seek the core road accident in it.Since algorithmic characteristic can come to add judgment of road accident loss and grade of severity by adding condition. Algorithmic can meet the need of specifications of traffic accident black spot checking, the most important thing is that the algorithm can find out the more accident road area at any position of the road.Rough collection theory is proposed by Poland mathematician Z.Pawlak in 1982, it is one kind of mathematics implement dealing with ambiguous and uncertainty problem. Its main characteristic lies in the information that need's any about data does not get ready for. The high-effect reduction algorithm is the foundation of data excavates and knowledge discovers applying rough collection. The paper mainly studies rough collection attribute reduction and decision-making rule reduction algorithm preferring and applies the algorithm on practice project. i.e., applying it on reason analyze of traffic accident, in addition, implementing the traffic accident analyze system.The paper is given out owing to the decision rule reduction algorithm preferring, adopt the judge , iteration crossing , regulation extraction etc. to operate the initial feature sets, making use of circulation iteration tactics to be accorded with consumer. Have been put into use coming to describe a decision table regulation form being connected , have quoted the equivalence kind coming to describe a decision table's characteristic collection and the concept respectively; And marriage partner in being a decision table has led into frequencies attributes, now that frequencies attributes not belonging to the condition attribute, belong to the attribute making policy neither , be only the number being used to record the identical marriage partner; The collection forming the once basic characteristic has agreed well with the value making use of every attribute condition choosing , defining deletion and intersect arithmetic on feature sets, forming by agreeing well with iteration to characteristic collection. By assigning regulation minimal support degree threshold value alpha and minimal confidence factor threshold value beta, accorded with the decision rule that the consumer demands can be got.Support degree and confidence factor threshold value choice have direct impact to regulation come into being. Very many regulation quantity meetings reduce algorithm efficiency, increase redundancy regulation; Regulation quantity stops being able to lose valuable information very much. The paper inducts the concept of expecting confidence factor and throwing up and scattering degree to evaluate the regulation validity. Expecting confidence factor describes the decision-making preferring information probability appearing under condition in random, scattering degree is the ratio between regulation confidence factor degree and expecting confidence factor, describing that how old conditional putting in produces in regulation to the regulation coming into being. Making use of above-mentioned theory and achievement to design black spots checks software. The software analyses the past previous years data in the Chang Chun City road accident database, checks the more accident area and finds out accident reason and provides the auxiliary decision-making effect to traffic management department. The software includes four modules: accident coordinate entering, more accident checking, the data counting and decision analysis.Accident coordinate entering module mainly provide the data source to other three modules, it combines GIS(geographic information system) and uses the electron map to process road accident data by standardization, being used to convert initial more accident area into coordinate, making it be distinguished by electron map; more accident checking is used to analysis the accident condition on every road area and section, finding out an more accident position according to coordinate information, in result being coupled back returning to the electron map; the data counting module uses the mathematics counts method to calculate out accident necessary relationship among accident reason, the accident from, the road and environmental factor; The decision analysis module applies mainly the decision rule reduction algorithm preferring, from analyzing correctly with algorithm a decision rule can be got and we can find out the impacts produced by road traffic and environment factors.Aiming at road accident characteristics such as delivering because of nature , conceal , sending out suddenly etc, applying a data excavate theory to research road accident regularity and combining relating regulation, clustering, rough collection theory and concrete application to develop Chang Chun City road accident analysis system. The software turn to be able to analysis data in the Ministry of Public Security road accident data base, finding out the more accident area; Making use of a data excavate technology to construct road accident analysis model, further, discovering the latent accident black spot and hidden trouble. The result of the software running can provides auxiliary decision-making information for related department. Adopting proper prevention to black spot would cut down hidden accident and ensure state property and the people life and property safety.The software is universal useful, it can help to check and analysis accident black spot in corresponding traffic area provided apply different area GIS and traffic accident information in it.