Conclusions 110 

8.1.1. Statement on problems at traffic signals

At most of the signalised intersections in MDCs, traffic process occurs disorderly and ineffectively. Under traffic engineering aspect, three major problems at traffic signals in MDCs still exist, which are traffic safety, intersection layout design, and signal program design. Of these three problems, traffic safety is a consequence of the other two problems under the mixed traffic condition.

Therefore, this study focuses on solving the problems of intersection layout and signal program design including traffic signal control strategies in order to improve traffic safety and increase traffic flow quality.

8.1.2. Measures for solving problems 8.1.2.1. Applicability of RiLSA in MDCs

The analysis of RiLSA in chapter 3 confirmed that RiLSA could be applied to MDCs with some modifications of intersection layout, signal program elements as well as control strategies. These modifications mainly based on the criteria of traffic safety and traffic flow quality.

8.1.2.2. Intersection layouts

The intersection layouts have to consider all types of vehicles: motorised traffic (car and motorcycle), public transport (buses), pedestrians, and cyclists. Measures for the layout of pedestrian and cyclist facilities can be taken from RiLSA, therein.

For motorised traffic, going from the observations at intersections, during the red time the motorcycle drivers always try to get in front of cars. Therefore, the main ideal in designing the intersection layout is to give the motorcycle riders an opportunity to get in front of cars. This solution utilized the red time to make traffic process stable on the approach before the green time starts. Hereby, it improves traffic safety, increases traffic flow quality and enhances capacity of intersections, as well, this solution of the intersection layout also allows to apply the leading green time for motorcycles waiting ahead in order to reduce the number of conflicts between motorcycles and pedestrians. However, on the exclusive right-turning lanes as well as right- turning carriageways, it is not necessary to use this solution because right-turning movements are not as critical as left-turning or go-through movements.

For public transport, if the road has more than 2 lanes and a high frequency of buses, an exclusive lane for buses should be established. Hereby, signalisation of buses can be taken from RiLSA.

However, it is noted that all the existing intersection layouts have been designed for car-oriented traffic already. Therefore, the adjustments of intersection layouts should only be implemented by marking. Only exceptional cases, the constructional measures will be used.

Chapter 8: Conclusions and Recommendations

8.1.2.3. Signal program elements

• Saturation flow

Unlike the conventional approach on saturation flow under the mixed traffic condition by using the equivalent number converting motorcycles into passenger car unit, this study gave the result of saturation flow depending on traffic volume of cars and motorcycles as well as homogeneous saturation flow of car and motorcycle. This is completely reasonable, because:

- firstly, saturation flow depends on the composition of traffic flow (proportion of motorcycles and cars in the traffic flow). Different proportions of motorcycles will give different results of saturation flow.

- secondly, saturation flow depends on “degree of mixture” of traffic. The more mixture of traffic occurs, the lower is the saturation flow (in this study, the adjustment factor f represents for degree of mixture of traffic).

In general, at each intersection, it is sufficient to determine saturation flow in peak hours, normal hours, and off-peak hours to develop the signal programs.

• Cycle time

From the concept of saturation flow, this study provided the new formulas to calculate the optimal delay cycle time and the minimum necessary cycle time. The result shows that the higher the degree of mixture of traffic, the longer is the cycle time. And the more motorcycle traffic volume occurs, the longer cycle time is.

However, the maximum cycle time should be 120 (150) s. Otherwise, the waiting time will be too long.

• Green time

Depending on the method for calculating the cycle time, the green time will be correspondingly calculated. This study provides the formula to calculate the green time according to the method for calculating the minimum necessary cycle time.

The minimum green time in the fixed-time signal program is 10 s to ensure that all motorcycles waiting in the head-start area can be discharged during this period of time.

• Amber time

Two major elements affecting the amber time are the speed limit at traffic signals, and the deceleration rate of vehicles. Therefore, the amber time is determined depending on these two elements that were stipulated in the traffic law of each country.

In MDCs, the amber time 3 s is determined for the case of intersections inside densely populated areas, and 4 s for the case of intersections outside densely populated areas.

Chapter 8: Conclusions and Recommendations

In Vietnam, the speed limit for motorcycles inside densely populated areas is 40 km/h, therefore the clearing speed must be lower than that in RiLSA. Depending on cases of the go-through vehicle or the turning vehicle, the clearing speed is 8 m/s or 5 m/s, respectively.

8.1.2.4. Control strategies in MDCs

All the control strategies related to counting traffic volume usually are not possible in MDCs because of motorcycle traffic. Therefore, only the macroscopic control level A1 (time-dependent signal program selection) could be applied to MDCs.

Regarding the microscopic control level, fixed-time signal control and traffic-actuated control can be applied to MDCs, in which traffic-actuated signal control includes green time request, time headway, demand phase, and time-offset adjustment.

According to the German control strategies, all the traffic-actuated control strategies above use only the detective function for vehicles. Therefore, for two-wheel vehicles (motorcycles), it is necessary to have measures to enhance the detective ability of the detectors. The proposed measure is rotating the inductive loops to an angle of from 30 to 45 degrees.

8.1.3. Draft of Guidelines for Traffic Signals in MDCs

Based on the German Guidelines for Traffic Signals RiLSA edition 2009 and some necessary modifications above, the first draft of Guidelines for Traffic Signals in MDCs was compiled. However, it should be noted that this draft is basically written based on the Vietnamese traffic law as well as on the traffic data collection in Vietnam, where many typical motorcycle dependent cities exist.