Development of a saturation flow and delay models for signalized intersection in Dhaka city

Abstract

Tramc in Bangladesh consists of both motorized and non-motorized vehicles, as in many other developing countries. The static and dynamic characteristics of the different vehicle vary widely even within the same class. Also, the lack of lane discipline and unrestricted mixing of the various types of vehicle in the same right of way makes the traffic stream heterogeneous in nature, The equation of delays and optimization of the signal timing which involve in predicting passenger car unit (PCU) and saturation flow, developed lor usc in developed countries, do not take into account the non-lane based traffic conditions prevalent in Bangladesh, However. the increasing number of vehicles and limited new road construction makes the correct prediction of these parameters important so as to make the most efficient use of existing road system by providing beller traffic operation and control. Intersections arc nodal points in transportation network and their efficiency of operation greatly innuences the entire network perlormance. Capacity analysis of signalized intersections basically consists of estimating saturation flow and delay, Pretimed signals are most commonly used in developing countries. Present study deals with development of saturation now and delay models for pretimed signalized intersections with reference to non-lane based traffic condition prevailing in Bangladesh. To carry out analysis trafJic data were collected at live signalized intersections in Dhaka city. Video camera was used to collect field data because it provides permanent record 01' data with minimum manpower. Recorded lilm was replayed in the laboratory and required information for saturation now and delay estimation was retrieved. Passenger car unit (PCU) values lor each vehicle is found out using synchronous ,; regression technique and a range of site-speei lic PCU values were obtained. The saturation flow lor each survey approach was calculated using the average PCU values and multiple linear regression techniques were then used to derive predictive saturation flow models. The saturation flow values for lilll approach were regressed against several intersection characteristics such as signal timing, ex, right turn, approach width and exit width. This modcl shows good correlation with ficld observed saturation flow and can bc used to estimate saturation flow in non-lane based traffic condition. Field delay for each approach is calculated based on HCM 2000 guidelines. Theoretical delay is estimated by various delay models and compared with lield delay. Based on results obtained fi'om analysis, HCM 2000 delay model has bcen sclected to modify. It has been observed that HCM 2000 delay model consistcntly over estimate delay at degree of saturation more than 1.0. It has been suggcsted li'om the analysis that theoretical incremcntal delay (due to random arrival and over saturated queues) in HCM 2000 delay model be reduced by 70 'X, to beller reflect field conditions. Recommended delay model shows good correlation for the present case but needs to be refined f1ll1her to become suitable for all range of input variables. Finally, simulation technique has been adopted to estimate number of stops incurred by each vehicle lacing an over saturated signalized approach. The research effort also deals with the generation of a stop estimation analytical model for over-saturated conditions through the establishment of an upper bound for the number of vehicle stops at oversaturated conditions using observed simulated data.