California PATH task order 4143
Task 2: Hybrid Data Fusion for ATMIS Applications
Researchers: Lianyu Chu
Time:  2001.8 – 2003.6
Project status: on-going

Project Summary:

This project will investigate how to fuse probe vehicle data, historical and real-time loop data in order to provide travel time information for real-time traffic control and management. PARAMICS simulation is used as an evaluation tool to evaluate the proposed artificial intelligence based data fusion method.

Project Description:

This project includes the following sub-tasks:

1. Study on the section travel time estimation methods based on mainline loop detector station data. The section is defined as the freeway segment between two adjacent mainline loop detector stations.

2. Study on the best aggregation cycle for travel time estimation. Sensitivity analysis is performed based on probe vehicle data and loop data.

3. Method for estimating travel time data from probe vehicle data

(1)  Based on the number of samples within a cycle

(2)  Considering the errors caused by the method of data collection (GPS, DSRC, Cellular phone, loop and video based signature…)

4. Study on the accuracy of travel time estimation from each data source, i.e. loop detector data and probe vehicle.

(1)  Relationship between percentage of probe vehicles and reliability

(2)  Fuzzy logic method (inputs: time-of-day, volume, occupancy)

5. Study and develop the algorithm to fuse loop detector data and probe vehicle data.

6. Evaluate the proposed fusion algorithm based on a case study.