40 likes | 123 Views
A real world problem: Predicting travel time from Lahti to Heinola A research was carried out on main road 4 between points A (Lahti) and D (Heinola) in Southern Finland. The average daily summertime traffic on this 28 km section is about
E N D
A real world problem: Predicting travel time from Lahti to Heinola A research was carried out on main road 4 between points A (Lahti) and D (Heinola) in Southern Finland. The average daily summertime traffic on this 28 km section is about 15100 vehicles per day. The study section AD is divided into three sub-sections AB, BC and CD with camera stations approximately equally distributed over link AD length and equipped with an automatic travel time monitoring system. The system is based on an artificial vision and neural network application, which automatically reads license plates. Moreover, there is an inductive loop detector on station C gathering information on traffic volumes and point speeds. A variable message sign (VMS) at point A gives upper and lower bounds of a forecast about the travel to the point D. The prediction classes are 20-25 min, 25-30 min, 30-40 min, 40-50 min and above 50 min. Travel time from point A to point D is regarded as congested if it is above 25 min. Point C Heinola Point D Point B Lahti Point A It will take 20 - 25 minutes to get to Heinola
The data was collected during the summer 2000 and processed into the following seven columns form Typical data sets contained 3700 - 19 000 rows corresponding to 40.000 - 150.000 vehicles
Based on GUHA analysis, the rule base of a Many-valued Similarity - inference system is the following (unit of measure is min) IF AD >= 23 AND AB+BC >= 17.5 AND 23 <= AB THEN PR = > 50 min IF AD >= 23 AND AB+BC >= 17.5 AND 14 =< AB < 23 THEN PR = 40 - 50 min IF AD >= 23 AND AB+BC >= 17.5 AND 10.5 <= AB < 14 THEN PR = 30 - 40 min IF AD >= 23 AND AB+BC >= 17.5 AND 5.58 <= AB <10.5 THEN PR = 30 - 40 min IF AB+BC+CD >= 21.25 AND BC>=6.3 THEN PR = 25 - 30 min IF AD >= 35 AND BC>=6.3 THEN PR = 25 - 30 min ELSE PR = normal (< 25 min) Results among a typical data:
The present model predicts the travel times right in 95,4% of all cases Among the congested cases the figure is 32,9% By GUHA-Similarity model, the travel times are predicted right in 98,8% of all cases. Among the congested cases the figure is 78,2%