Study On Freeway Traffic State Estimation Method Considering Speed Difference Characteristics

Estimate highway traffic state accurately which is the foundation of improving travel efficiency and the protection of traffic safety. The traffic parameters which are utilized in state estimation algorithm can reflect the traffic state macroscopically and medially, which may cause that there is a gap between the estimated results and the actual state for the reason of ignoring the difference between the individual vehicles' traveling state in the real traffic environment and the lack of comprehensive traffic operation state of traffic flow. Thus, to estimate the traffic state comprehensively and accurately, we need to take the consideration of the vehicle state difference and take the speed difference as the object of study, which has the theoretical and practical significance for improving the effect of highway state.This paper will take the basic section of highway as the study object and take the consideration of speed difference. With studying the distribution rule of speed difference and the relationship with traffic state, the paper proposes the traffic state estimation methods in both road and location of the freeway to promote the accuracy of estimation method and make the result is more close to the feeling of people. The main contents are:(1) First, summarize the manner of definition of speed difference on the basis of statistical theory and propose the definition of relative speed difference. Then research the variation of speed difference under the condition of different speed limits and the carts proportion. Finally, research the relationship between speed difference and traffic density with the same average speed through the exploring the speed difference and distribution rule of traffic parameters in the free flow and congested traffic state on the basis of fundamental diagram of traffic flow(2) Model the relation modeling among speed difference, average speed and density based on the safety distance linear model which overcomes the limitation of traditional state estimation of section of freeway. Establish the traffic state estimation standard combining the speed difference and average speed on the basis of traditional ways which take the average speed as the absolute estimation indicators, and then estimate the traffic state. The experiments demonstrate the effectiveness of this method.(3) There is a certain one-sidedness taking macroscopic parameters as state estimation features about the traditional location state estimation based on fuzzy C-Means clustering. Thereby determine the different features' weights in model by introducing the speed difference index and using the method of ReliefF according to the different results contribution of different features. And then improved the state estimation method by artificial bee colony algorithm, which is not sensitive to the initial cluster centers. Finally, verify the validity of the method proposed by this paper by experimental analysis.Above all, this paper proposed a state estimation method of freeway considering the speed difference. The experiment results show that this method has a certain advantages about the subjective and objective indicators and matches with the traffic participants' intuitive feelings.