Research On Early Warning Of Highway Traffic Abnormal State

With the rapid development of the expressway,the early warning of expressway traffic is getting more and more attention.The timely prediction and early warning of the abnormal state of the expressway involves the safe operation and construction of the expressway and plays an important role in the smoothness of the expressway.Relatively closed sections of highways are characterized by fast speed and high flow rate,and there are many types of associated influencing factors.Therefore,to complete the forecast and early warning in a short period of time requires the entire early warning system to be fast,accurate,and stable.The main research work of this paper is as follows:First of all,according to the relevant data of highway traffic situation,this paper adopts a traffic attribute reduction method based on rough sets,and combines the data of freeway traffic situation-related fields to reduce dimension and make it more convenient for subsequent processing calculations.The research focuses on the study of reduction methods and the identification of as few attribute nuclear objects as possible after the reduction.The emphasis is on the correspondence between the abnormal state and the traffic situation data and the analysis of the dependence degree of the influence degree.Then,in the aspect of highway traffic situation forecasting,this paper will first use the short-term forecasting method to predict the reduced attribute nuclear,and draw its development trend.The purpose is to efficiently,accurately,and reasonably predict the traffic conditions of the expressway in the future,so as to better serve the highway warning service.Based on the pretreatment of traffic data,the traditional traffic parameter prediction methods are compared and analyzed.The idea of improved combination forecasting is put forward,and an improved method for predicting short-term traffic flow parameters is proposed.Finally,through the establishment of the status discriminating module and the graded early warning module,it provides preparation and support for the analysis and study of the abnormal state of the highway.Combining traffic situation reduction and forecasting values as an early warning,the main factors related to the characteristics of traffic entities in the traffic situation,including analysis of traffic flow,speed,and other data.Co-factors include weather conditions,wind levels,and temperatures.The state of the expressway is evaluated by K-fuzzy clustering,and the main reason for the abnormality is analyzed according to the maximum degree of membership.