Prediction of Vessel Traffic Accident Based on Chaotic Theory

Abstract

It is well known that the vessel traffic accident is a quite complex event induced by many kinds of factors, which can be concluded as persons, ships, and environment. As a result of interaction and intercoupling of these factors, the whole system has highly nonlinear characteristics. Due to the limitation of traditional linear prediction, it is significant to put the chaotic theory into the vessel traffic accident prediction. As a new attempt this paper at first analyzes chaotic characteristics appeared in vessel traffic accident, and then presents the chaotic adaptive prediction model based on wavelet de-noising. In theory this model is suitable for the vessel traffic accident prediction whose history record contains only small data sets but with much noise. In practice the following simulation results on MATLAB software platform show the effectiveness of the model described which has high prediction accuracy and can meet the actual need.