基于大规模实景路况与人工智能的交通平台——以实景路况为基础的多层级分布式AI模型与传统建模相结合的交通系统

本文提出以实景路况为基础的多维度、多层级分布式AI模型与传统建模相结合的实景路况交通平台解决道路资源合理化分配以及调控的问题。大规模动态实景路况可以实时给出更高维度的路况信息,打破现有算法滞后性带来的局限;分布式AI模型可以在大幅度降低模型复杂度的同时,降低系统误判带来的“蝴蝶效应”式影响,从而大幅度提高系统的稳定性;多层级AI模型可以在街道(乡镇)、县级市(县、区)、地级市、省(直辖市、自治区)以至于国家层级进行调控,打破AI模型在交通中难以人为调控的难题。本文研究结合传统模型可以在AI模型发生误判时,大幅度降低不良影响,及时止损,并讨论了以该模型建立交通平台可衍生的应用场景。

In this paper,a traffi c platform with real road conditions,which combines a multi-dimensional,multi-level distributed AI model based on real road conditions and the traditional modeling,was proposed to solve the problems of rational allocation and regulation of road resources.Large-scale dynamic real road conditions can provide higher-dimensional traffi c information in real time,which breaks the limitations caused by the lag in the existing algorithms.The distributed artificial intelligence(AI)model can greatly reduce the complexity of the model while reducing the impact of the"butterfl y effect"caused by the misjudgment of the system,thereby greatly improves the stability of the system.The multi-level AI model can be controlled at the town,county,city,province and even country levels,breaking the diffi cult problem of manually controlling the AI model in the traffi c management.This research fi nding,combining with the traditional models,can greatly reduce adverse effects and timely stop the losses when AI models make mistakes.The paper also discussed the application scenarios that could be derived by using this model to establish the traffi c platform.