601 / 2019-03-01 22:21:56

A Novel Urban Road Management System Based On Data Mining

data mining,time series,SVR,ARIMA,urban road management

全体主题 > Urban Management Systems

As a part of urban public space, municipal roads are closely related to the daily activities of urban residents. It is also city’s management duty to keep the conditions of urban roads well. With the rapid development of big data and the continuous exploration and practice of smart cities, how to effectively use data mining techniques to analyze massive historical data to provide helpful guidance is an important and valuable topic.
After the in-depth study of historical data on street cases of Ningbo, this paper proposes a prediction model for the number of street cases. After reviewing all relevant influencing factors, this model firstly used combined SVR model to fit the short-term fluctuation characteristics of the dataset. And then we used ARIMA model to find the long-term trends hidden in the residual sequences obtained from the first step. Finally, we added the values predicted by the two models to get the final prediction. The experimental results indicate that the results can match the real data well. So based on this model, we developed a prototype of a novel urban road management system(NURMS), which realized some useful functions such as inquiry of daily cases, prediction of the number of daily cases and statistical analysis of historical data.