Short-term Traffic Flow Uncertainty Forecast Based On Fuzzy Theory

The control and induction of traffic flow needs to be based on accurate traffic flow prediction.In the process of traffic flow prediction,the influence of various factors will lead to deviation of the prediction results,which is defined as the uncertainty of traffic flow prediction.Studying the uncertainty of traffic flow prediction can provide decision support for traffic management and control.Therefore,the text makes the following research on the construction of the traffic flow uncertainty prediction model.Firstly,it is very important to deeply analyze the development status of traffic flow prediction and fuzzy theory at home and abroad.Traffic flow prediction is divided into two types: mean prediction and uncertainty prediction.Compared with the well-developed mean prediction method,research on uncertainty prediction is still scarce.However,the fuzzy time series method which is more mature in fuzzy prediction theory can only realize point prediction,and the research based on fuzzy theory to realize uncertainty prediction is very scarce.Therefore,this thesis aims to introduce fuzzy theory into traffic flow uncertainty prediction,and construct a short-term traffic flow uncertainty prediction model based on fuzzy theory to realize traffic flow interval prediction.Secondly,in this paper,the original traffic flow data preprocessing methods including data collection,missing data supplementation and error data repair are deeply analyzed.On this basis,the data used in this paper is described,and the error data and missing data are corrected and supplemented.And according to the existing traffic flow time collection degree selection method,the data is collected and processed according to the aggregation degree of 10 min,15min and 30 min.Finally,based on the in-depth study of fuzzy theory and fuzzy time series theory,this paper constructs a short-term traffic flow uncertainty prediction model based on fuzzy theory with "fuzzification" and "defuzzification" as the core idea.The corresponding predictive performance evaluation indicators are given in this paper,namely the kick off percentage and the width flow ratio.And the model is applied to the actual traffic flow uncertainty prediction,and compared with the existing SARIMA+GARCH model.Research shows that the kick off percentage of this interval is zero,that is,there is no case where the actual value point falls outside the interval.Under the premise of ensuring that the kick off percentage is zero,the interval width flow ratio remains at a small level.In summary,the traffic flow uncertainty prediction can be realized by the model established in this paper,and the predicted interval has practical utilization value.