טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
M.Sc Thesis
M.Sc StudentJack Haddad
SubjectA Model for Estimating Traffic Performance of Paired
Signalised Intersections
DepartmentDepartment of Civil and Environmental Engineering
Supervisor Professor Mahalel David


Abstract

    As traffic becomes more congested in urban areas, an operational analysis of Paired Signalized Intersections (PSI) becomes more complex.  Two signalized intersections are defined as PSI when queue spillback and intersection blockage occur.  PSI have their own consideration regarding operational characteristics that are different from isolated intersections and or coordinated networks. Typical examples of PSI are staggered intersections and diamond interchanges.

    The movements that are most affected by the short distance between the two PSI intersections are those that pass through both intersections.  Other movements can also be affected whenever the Upper Stream Intersection (USI) is blocked by a queue that starts at the Down Stream Intersection (DSI).  Because of the possible intersection blockage, only part of the green period is utilized by the saturation flows.

    The present study examines and expands an analytic model called PICADA. The model determines the influence of the downstream queue on the discharge profile at USI.  The CP--Clear Period--is defined as the time duration of the cycle when the blocking queue does not exist.  The modified effective green is the time during which the upper stream movement can be discharged; this occurs when the effective green period overlaps the CP.

    Special care is given to flared lanes at the downstream intersection.  Probabilistic expressions are developed for estimating the chances of queue spillback from short turn lanes to through lanes.  Capacities of the upper stream movement are calculated for various DSI situations.

The model has been programmed and integrated into a computer software called INBAR, which is capable of planning and evaluating signal programs.

Intersection capacities were calculated by the model for various signal programs and compared to capacities resulting from simulation runs.  A comparison of the two sets of results showed only minor deviations.

    The effects of offset, green period and cycle time have been demonstrated through sensitivity analyses. The capacity of USI movements varies significantly in relation to offset values. For the same timing plans, capacity may be increased fourfold just by changing the offset. It is essential to change offset values according to volumes and to choose a value that best reflects the planning objectives.