퍼지 추론을 이용한 교차로에서의 사고예측기법에 관한 연구
(A) Study on the analysis of the predicting model of traffic accidents at intersections by using fuzzy inference
퍼지추론 교차로사고예측기법 도시공학;
- 원문 URL
The accident at the intersection is due to many combined factors, drivers, cars, environmental traffic conditions, traffic facilities and so on. Presently major traffic accident assesment is considered with exact probability value quantitative value. So, In many factors causing the intersection accident, the consideration of unquantifiable or Fuzzy characteristic factor is deficient. In this paper, we applied the manner of Fuzzy Approximate Reasoning in consideration of vague Fuzzy variable and made an estimate the accident event at the intersection and construct a model. We made a Fuzzy Rule and a model based on Fuzzy Membership Function through the question survey. We made sixty-four intersections the object of all intersection in Jin-Ju and compared the actual intersection accident event with the presumed volume of the model. In the result that We measured the actual accident event and presumed accident event come under each rule, we can know that appropriateness of Fuzzy variable is reasonable. Besides, In the study there is a tendency that intersection accidents increase when the volume of is heavy and the speed is fast and the brightness is feeble. When it is tested with statistical analysis method, it illustrates that R^(2) is 0.971, MAE is 0.08708 and MSE is 0.18204. It shows that the prediction of Fuzzy approximate Reasoning model is excellent in estimating Traffic Accident at Intersections.