Pedestrian Traffic Flow Forecasting Method Based On Hadoop Big Data Platform

Over the last decade, traffic jam has intensified around cities over the world. With the increasing urban population, the purchase rate of private transportation rises gradually, whereas the penetration of public transportation declines all the time. Many countries have developed big data as national strategy since 2011 and regard the transportation problem as a starting point of big data strategy. Thus, it is of great significance to solve the transportation problem by Hadoop. So far, the urban transportation efficiency is very low. As we know, traffic jam arises with the increasing urban population. More importantly, the backward transportation system exacerbates it. A huge amount of data returned from the transportation testing measurement cannot be effectively analyzed. As a result, we cannot get access to real time statistical analysis and the transportation problem cannot be solved effectively by reliable feedback data. On the other hand, the transportation data is only available for the governmental officials and that makes it impossible for the whole society to brainstorm and build up a more effective traffic management platform. Consequently, the transportation problem cannot be solved in a proper way. In a word, it is very important to combine a huge mass of data offered by government and state-owned enterprises and do research on Hadoop technology to better the situation of traffic jam.The main contributions of this paper are as follows:(1)We explore various methods to improve traffic jam through big data platform. Since the smart transportation management system is very complicated and multiple methods are used to collect information of users and vehicles, there are various methods to better traffic jam, each of which has respective advantages in different scenarios. For example, we can apply big data technology to smart control over traffic signal lights. And then we can obtain optimal control method to adjust working time of all the traffic signal lights in the city by collection of real time vehicle data and distributed algorithms of big data platform.(2) We build up a pedestrian traffic flow forecasting platform based on Hadoop big data technology. The construction of Hadoop platform is the foundation of statistical analysis of related data. Thus, it is very critical to build up, maintain and utilize Hadoop big data platform in a flexible way. With the increasing family members of Hadoop platform, the method of data processing of Hadoop becomes richer and richer gradually and the capability to solve traffic jam also turns out to be stronger and stronger.(3)Hadoop platform obtains the trajectory of a larger number of users by analyzing the data from mobile operators and then predicts pedestrian traffic flow. Through simulations, we can proof the data-processing capability of Hadoop platform. By analyzing the trajectory of a user, we can pre-judge the most likely mobile trajectory of it, which is helpful in judging the most likely blocked area in advance and then making timely adjustments to reduce the occurrence of traffic jam to the most extent.