RETRACTED: Optimal logistics transportation and route planning based on fpga processor real-time system and machine learning

Abstract

The shared bus's development requirements, relief of traffic congestion in urban areas, and improved utilization of the road resources providing the transport mode of the excellent user experience neotype are very urgent. To predict precise travel needs, the key for planning a dynamic routing lie a lie of the shared bus implementation. However, the shared bus data's sparse and high volatility will require a lot of resistance to predict travel accurately. Based on the user experience, very different from the traditional public transportation that is far more challenging to the relatively high number of optimization goals is because passengers of the shared bus route planning and shared bus route planning. This article, based on the shared bus data from different audiences sources, travel demand prediction and dynamic route planning in "the last mile", and a two-step process that consists of the shared bus dynamic routing (sub-bus), proposed and your scene. First of all, such traffic, time, week, location, and five of the prediction function such as a bus, to analyze residents' travel behavior to prepare the travel demand based on them precisely the machine learning model used to predict. Secondly, dynamically and predict the results of multiple operations bus optimal routing, designed to generate a fixed based on the shared bus destinations' operating characteristics, a dynamic programming algorithm wants below. Several experiments, based on the shared subway shuttle bus of evidence that people of the data and the reality has been purchased, the sub-bus is better than the method of dynamic route planning, etc. for the scene of such a "last mile".