Smart Signal Control Strategy

Primary Traffic Data Collection

Primary traffic data like traffic volume, travel time, queue length and speed, etc. were collected on all three selected junctions. Traffic data was collected for 16- hours in a day starting from 6am covering morning, afternoon and evening peak hours. Video graphics techniques was used to collect classified traffic volume and queue lengths whereas for speed and travel time, floating car methods was adopted. All the intersections were working on fixed time mode where cycle time and green time varies from peak to off peak hours in the day. As a result of primary traffic data, it is observed that during evening peak hour between 17.15-18.15, traffic on study corridor accounts for 9.3% of total 16hours traffic. It was also identified that average speed on the study corridor was 29kmph and average journey delay was 88 seconds.

Integration with local controller

The local signal controller was operating with fixed time mode and had very limited functionality for dynamic traffic signals. An intelligent small device called Centris (developed by the Turkish company ISSD) was integrated with the local signal controller, thereby enabling it for adaptive traffic signal control. The Epics algorithm which is running inside Centris will optimize the traffic flow second by second with respect to the traffic demand collected from detection cameras. The Centris will act as a master controller whereas local controller acts as a slave unit. PTV Epics analyses traffic conditions locally and determines the best control option for the single node and Centris will trigger the local controller to switch traffic lights accordingly.

Installation of Detection Camera

On each direction of traffic, a camera was placed at 50-60m away from the stop line to measure downstream traffic at the height of 6-8m from the ground level. The

detected traffic data were transmitted to Centris which was located inside the signal controller cabinet via network cable. In non-lane based traffic, success ratio of 80-90% was observed by camera based detection. The optimization algorithm PTV Epics collects the detector data inside Centris from camera to produce new signal control strategy for a local junction.

Traffic Management Center

All three junctions were monitored centrally inside traffic management center for operation and maintenance. PTV Balance located centrally in traffic management center collects the data from individual signal controller and determines OD pattern on the corridor and optimizes the corridor every five minutes by sending signal plan to PTV Epics to optimize the traffic flow at junction level.

Post Installation Results

After testing and calibration of the system on ground, the traffic survey was carried for post installation to evaluate the improvement of new signal strategy with dynamic traffic signal. Measurement includes traffic flow, queue length, speed, delay etc. It is observed that overall network traffic flow was increased by 6% during the same peak hour as before with fixed signal plan. Pre-and Post traffic study was validated with statistical student T-test and it was found that at 95% confident interval, both flows did have significant differences. The New signal control strategy showed significant improvement in different traffic performance indicator. The travel time, number of stops and delay has been reduced by around 25%-45%. Overall network speed has been increased from 28kmph to 33.7kmph during a peak hour with an improvement of 20%.