Detection Of Traffic Congestion And Its Application Based On Data Mining Technology

The municipal transportation crowded question is the universal question. Looking from the worldwide scale, except minority several cities, because the developed public transportation system as well as limits to the compact car transportation development or is the urban population and the vehicles are very originally few, therefore traffic congestion phenomenon not very serious outside, as for the whole, the world majority national varying degree is suffering the traffic congestion the puzzle. Because the transportation information supplies insufficiency and the management measure are disadvantageous, enables the existing traffic equipment to display the proper ability, has created the traffic order confusion and the transportation efficiency is low, has brought the huge hindrance for the entire social development. In the past the traditional solution traffic congestion the means were expand the existing road or construct the new path. But along with the construction land-levying and the roadbuilding expense's unceasing growth, uses this conventional routes solution traffic congestion the difficulty to be getting bigger and bigger. Moreover although assignment road capacity may alleviate the traffic congestion, but it has the economical effect rate difference, affects the journey time the shortcoming, it can be said that is also the method which an item not goings to the roots of the problem. At present, the recognition avoids the traffic congestion the method is the transportation passes leads. Then must rely on the traffic congestion the examination recognition.The traffic congestion the examination and the transportation induction information's issue depends upon the transportation monitor technology, the computer information processing and management technology and the communication realizes together, because the present this domain computer information processing and management technology is quite unitary, therefore causes the detector to establish the position limitation to be big, event detection rate low status numerous question.This article goal is through uses the data mining technology construction monitor data the inner link, establishes the monitor area the crowded model, like this ceased to the sole examination data dependence, avoided the detector establishing the position the limitation, reduced the construction cost; Simultaneously enhanced the examination rate, the accurate event will monitor the information issue for the transportation participant, helps them to avoid the saturated road section reduction traffic jam ahead of time the occurrence.Because the traffic congestion will bring the economy, the security, the environmental pollution and so on various question, will therefore solve it to have and the vital significance to the urban development.The data mining is from massive, incomplete, has the noise fuzzily, in the stochastic data, withdraws the concealment, the people does not know beforehand, but is also the latent useful information and the knowledge process. The current internationally most influential demonstration studies the method to regard J.R.Quinlan ID3. Its predecessor is CLS. The CLS work process is, first discovers most has the distinction strength factor, divides into many subsets the data, each subset chooses most has the distinction strength factor to carry on the division, has carried on until all subsets only to contain the identical type the data. Finally obtains a decision tree, may use it to come to carry on the analysis to the new type example.The J.R.Quinlan work is mainly introduces in the information theory the information mutually, is called the information gain it, takes the characteristic distinction ability the measure, to define the gain, must draw support from the information theory a concept - - entropy. The entropy may survey the attribute information content. Finally inlays the achievements method in an iterative outer covering. This article is receives this method the inspiration, processing utilizes in the traffic congestion examination it to the data, the construction decision tree, the optimized processing result, carries on the intelligent recognition to support stops up.After the experiment, proved that this method the recognition rate of accuracy is very high regarding the traffic congestion, the rate of false alarm is very low, is one very feasible method. Prompt issued the very good support function accurately for the transportation induction information.Along with the economical development, municipal transportation construction investment's enlargement, path detector's quantity will be getting more and more, induces the information the issue more dependences in the automatic detection system, can the simultaneous working many detector's information be this system's future development direction.