Exploration and Research on Intelligent Traffic Control System based on Cloud Computing

Abstract

As an important part of the intelligent transportation system, the short-term traffic flow forecasting is always one of hot research. The short-time traffic flow prediction based on MapReduce technology is put forward in the paper from the view of improving practical prediction algorithm of cloud computing platform and in combination with the technical advantage in mass data storage and massively parallel real-time processing. Improve the computational efficiency of the short-time traffic flow forecasting under the premise of ensuring the accuracy of prediction so as to enhancing the practical of the prediction algorithm. Simulation results show that the problems of long training time of the early network weights in BP neural network algorithm and of low efficiency for matching the searching history database in K nearest neighbor nonparametric regression algorithm pattern. And, it can effectively improve the prediction of the practicability of the algorithm. Interface Cloud computing is a new computing patterns developed in recent years. It integrates all software, hardware and information resources based on the existing internet to form new system architecture to provide services and to provide intelligent information service as low cost and high efficiency needed by people at all levels(1-2). The internet resources service provided by cloud computing is called cloud. Cloud computing has become a new industry with a huge potential and broad market prospects recognized in the world. At present, investment on cloud computing has been carried out around China(3-4). Some domestic companies as Baidu, Alibaba, Grand and Telecommunication Operators are also building their own cloud computing platform as the company's data processing and storage platform and commercial application(5). These commercial companies have brought new hope for promotion of cloud computing technology development, especially the complex and changeable traffic system area.