Construction and Application of Trajectory Data Analysis Model Based on Big Data and Stochastic Gradient Descent Algorithm

Abstract

This paper studies the model construction of computing and storage resource management system framework based on Hadoop and the implementation of trajectory data analysis function under big data. Relying on the cloud platform infrastructure, in order to support the rapid data growth and massive data processing needs, it provides a mixed storage and analysis platform for structured and unstructured data, and uses big data technology to build a highly scalable and distributed data processing framework. The distributed computation, overall frame model of the memory system, and function module have been built with the aim of constructing the system in consideration. Second, by using Hadoop to preprocess the original data and concentrating on the data hierarchical design model and key technology analysis of big data systems, the design model, functional modules, technological solutions, and SGD algorithm are suggested, along with the detailed implementation procedure. Lastly, by merging the data of running vehicles, the system accomplishes the data analysis of vehicle trajectory, empty and load cars, and load and unload people.