Q. Dong

Pages: 125-134

Abstract
In order to overcome the problem of low prediction accuracy of traditional road traffic sign occlusion failure risk prediction method, a road traffic sign occlusion failure risk prediction method based on fuzzy Bayesian network was proposed. The paper analyzes the types of road traffic sign occlusion, and preliminarily clarifies the reasons and principles of failure of road traffic sign occlusion. Introducing fuzzy Bayesian network, calculate the risk training focus on the joint probability distribution function with the input and output, input risk calculated by using Bayesian algorithm corresponding to the maximum a posteriori probability and traffic signs of failure risk probability calculation results, the risk probability calculation results input to build a good shade failure risk prediction model, the access to keep out failure risk prediction results. Experimental results show that the proposed method can effectively obtain the failure time of road traffic sign occlusion, the probability of drivers missing sign information and the probability of sight occlusion. Therefore, the proposed method has higher prediction accuracy and reliability.
Keywords: fuzzy Bayesian network; road traffic signs; occlusion failure; risk prediction; joint probability distribution functions