A multi-source data fusion model for traffic flow prediction in smart cities

The emergence of problems such as increased urban traffic and transportation, traffic congestion, and road resource shortages prompted the city to prioritize the construction of intelligent transportation. Intelligent computing technology provides technical support for people’s smooth travel, as well as for the construction and development of the city. The article first delves into the basic theory of support vector machines, outlines a specific process for traffic prediction using these machines, and then suggests a method for preprocessing traffic data. This method primarily involves three steps: data collection, abnormal data elimination, and missing data recovery. Finally, it proposes a traffic flow prediction model that utilizes multi-source data from support vector machines. The experimental results demonstrate a higher degree of consistency between the prediction results and the actual results in the analysis of short-time traffic flow prediction on weekdays and weekends. Furthermore, the traffic flow prediction model based on Support Vector Machine, proposed in this paper, is capable of reliably predicting vehicular traffic flow in various severe weather environments, with a coefficient of parity below 0.018.