This paper proposes an application of max-pressure control for network-wide signal control at Bandung, Indonesia. The max-pressure approach is employed for a specific disturbed network system synthetic scenario, created with the aim to simulate spillback conditions which causing long congestion across road segments in real traffic conditions. The max-pressure controller is implemented for a network of six signalized intersections in PTV Vissim, a traffic micro-simulation platform. The validated model is generated before implementing in the Vissim traffic simulation. Three types of controller are studied: the currently implemented controller (fixed time controller), cycle-based max-pressure and slotted-based max-pressure. The simulation results show that max-pressure control is more powerful than the currently implemented technique in terms of the capability to avoid congestion by spreading vehicles to other road segments with respect to some events that can be seen as a disturbance.

The increase in traffic in cities world-wide has led to a need for better traffic management systems in urban networks. Despite the advances in technology for traffic data collection, the collected data are still suffering from significant issues, such as missing data, hence the need for data imputation methods. This paper explores the spatiotemporal probabilistic principal component analysis (PPCA) based data imputation method that utilizes traffic flow data from vehicle detectors and focuses specifically on detectors in urban networks as opposed to a freeway setting. In the urban context, detectors are in a complex network, separated by traffic lights, measuring different flow directions on different types of roads. Different constructions of a spatial network are compared, from a single detector to a neighborhood and a city-wide network. Experiments are conducted on data from 285 detectors in the urban network of Surabaya, Indonesia, with a case study on the Diponegoro neighborhood. Methods are tested against both point-wise and interval-wise missing data in various scenarios. Results show that a spatial network adds robustness to the system and the choice of the subset has an impact on the imputation error. Compared to a single detector, spatiotemporal PPCA is better suited for interval-wise errors and more robust against outliers and extreme missing data. Even in the case where an entire day of data is missing, the method is still able to impute data accurately relying on other vehicle detectors in the network.

The demand for transportation has increased significantly in recent decades in line with the increasing demand for passenger and freight mobility, especially in urban areas. One of the most negative impacts is the increasing level of traffic congestion. A possible short-term solution to solve this problem is to utilize a traffic control system. However, most traffic control systems still use classical control algorithms with the green phase sequence determined, based on a specific strategy. Studies have proven that this approach does not provide the expected congestion solution. In this paper, an adaptive traffic controller was developed that uses a reinforcement learning algorithm called deep Q-network (DQN). Since the DQN performance is determined by reward selection, an exponential reward function, based on the macroscopic fundamental diagram (MFD) of the distribution of vehicle density at intersections was considered. The action taken by the DQN is determining traffic phases, based on various rewards, ranging from pressure to adaptive loading of pressure and queue length. The reinforcement learning algorithm was then applied to the SUMO traffic simulation software to assess the effectiveness of the proposed strategy. The DQN-based control algorithm with the adaptive reward mechanism achieved the best performance with a vehicle throughput of 56,384 vehicles, followed by the classical and conventional control methods, such as Webster (50,366 vehicles), max-pressure (50,541 vehicles) and uniform (46,241 vehicles) traffic control. The significant increase in vehicle throughput achieved by the adaptive DQN-based control algorithm with an exponential reward mechanism means that the proposed traffic control could increase the area productivity, implying that the intersections could accommodate more vehicles so that the possibility of congestion was reduced. The algorithm performed remarkably in preventing congestion in a traffic network model of Central Jakarta as one of the world’s most congested cities. This result indicates that traffic control design using MFD as a performance measure can be a successful future direction in the development of reinforcement learning for traffic control systems.

The growth of vehicle mobility in the past decades and increased traffic complexity leads to a need for traffic management systems, especially in large-scale urban traffic networks. The erroneous data problems are common problems that affect traffic management systems. The traffic management systems also relied on traffic prediction particularly in traffic signal control and route guidance. This paper investigated probabilistic principal component analysis (PPCA) methods to impute missing traffic count data and predict future data. We also investigated the resulting principal components' significance in urban traffic analysis. These methods are applied to traffic count data from vehicle detectors in the urban network of Surabaya city, Indonesia. The results show that the PPCA-based data imputation method is able to impute missing data with imputation error under 20% WMAPE. The resulting principal components analysis demonstrates that 1 ^st principal component scores can be seen as a fundamental temporal pattern of the Surabaya urban network while the characteristic of the link can be derived from the 1 ^st principal component coefficient. We also demonstrate that 1 ^st principal component coefficient of the link might detect outliers or anomalies such as detector malfunction and unique temporal pattern. PPCA can also be used to predict future data based on observed data, but experiments show that even though the majority of the links can be predicted accurately, some links are having large errors that might be caused by different temporal patterns between future data and observed data.