Abstract
This study creates an adaptive data fusion procedure to represent the sequential forecast of accident duration. This procedure includes two Artificial Neural Network-based models. Model A is used to forecast the duration time at the instant of accident notification while Model B provides multi-period updates of duration time after the moment of accident notification. These two models together provide a sequential forecast of accident duration from the accident notification to the accident site clearance. With these two models, the estimated duration time can be provided by plugging in relevant traffic data as soon as an accident is being notified. Through the feature composition approach, the number of inputs can be decreased while the relevant traffic characteristics are preserved. This study shows very promising practical applicability of the proposed models in the Intelligent Transportation Systems context.
