Research On Key Technologies Of Intelligent Vehicle Control

In recent years,traffic safety problems have become increasingly prominent with the rapid development of transportation.Therefore,intelligent vehicle came into being using vehicle engineering, artificial intelligence,and other areas of the intersection of technology and integrated which is the development trend of future vehicles.In this paper,we have been made some positive and useful explorations in the field of intelligent vehicle control technology,aimed at supporting the development of automotive safety driving system.The control System is called "brain" of the Intelligent Vehicle.In this paper,an algorithm of clouds based on golden division and a "T-S-C-Y" reasoning model as the reasoning mechanism of the controller are proposed that use RBFNN to approach driving equation in the corresponding sub-areas of cloud.Regard the value under a cloud of sub-regional membership as the random confidence of the rule, and the overall random output of the whole nonlinear system is weighted as a average of the rule's random outputs.Autonomous driving and aided navigation are also the key technology of Intelligent Vehicle.This paper proposes the dynamic target position concept and the corresponding control mechanism.The lateral control,moving obstacles avoiding and lane changing and overtaking can be realized based on the dynamic target position concept.It can simulate ideally motion characteristic of vehicle and driver's behavior in realistic traffic situation.Intelligent vehicle control is not an isolated system,traffic signals control strategy will have a direct impact on intelligent vehicle control and thus affect the entire intelligent transportation system efficiency.By using STREAM algorithm to analyze the data flow of traffic junctions,we get the clustering results that can react the reality of different traffic conditions,and then make data mining and traffic flow trend forecasting.Finally,based on forecasting value of the data flow,we analyses the forecast traffic flow value of one day based on the cloud model and get a more flexible control strategy.