Traffic Signal Optimization

Streszczenie: Praca opisuje pomysł na system zarządzania ruchem pojazdów w dużej skali (np. miasta wielkości Warszawy i większe) bazujący na zastosowaniu symulacji komputerowej w modelu mikroskopowym, metaheurystyk (np. algorytmów genetycznych), obliczeń o wysokiej wydajności i sieci neuronowych. Rozważania teoretyczne sugerują, że system taki może działać lepiej niż współczesne systemy zarządzania ruchem. Wyniki eksperymentów są obiecujące i pokazują, że warto prowadzić prace badawcze w tym obszarze. Abstract: The article presents an idea for a traffic management system working efficiently in a scale of a large city (e.g., of the size of Warsaw or larger) based on microscopic traffic simulation, metaheuristics (e.g., genetic algorithms), high-performance computing and artificial neural networks. Theoretical deliberations suggest that such system may outperform existing traffic management systems, while initial experiments are encouraging and show that some research efforts should be put to investigate this possibility further. 1. Wprowadzenie Problem optymalnego zarządzania ruchem drogowym jest bardzo ważny z punktu widzenia gospodarki i efektywnego funkcjonowania współczesnych miast. W stosowanych już obecnie w Polsce i na świecie rozwiązaniach dużą rolę odgrywają nowoczesne technologie ITS. Powstają zaawansowane, zintegrowane systemy zarządzania ruchem, które potrafią działać w trybie adaptacyjnym, zbierają przy pomocy detektorów dane o ruchu drogowym, analizują je w czasie rzeczywistym i podejmują odpowiednie działania (np. rekonfiguracja sygnalizacji świetlnych),

In this paper, we focus on computing a consistent traffic signal configuration at each junction that optimizes multiple performance indices, i.e., multi-objective traffic signal control. The multi-objective function includes minimizing trip waiting time, total trip time, and junction waiting time. Moreover, the multi-objective function includes maximizing flow rate, satisfying green waves for platoons traveling in main roads, avoiding accidents especially in residential areas, and forcing vehicles to move within moderate speed range of minimum fuel consumption. In particular, we formulate our multi-objective traffic signal control as a Multi-Agent System (MAS). Traffic signal controllers have a distributed nature in which each traffic signal agent acts individually and possibly cooperatively in a MAS. In addition, agents act autonomously according to the current traffic situation without any human intervention. Thus, we develop a multi-agent multi-objective Reinforcement Learning (RL) traffic signal control framework that simulates the driver’s behavior (acceleration/deceleration) continuously in space and time dimensions. The proposed framework is based on a multi-objective sequential decision making process whose parameters are estimated based on the Bayesian interpretation of probability. Using this interpretation together with a novel adaptive cooperative exploration technique, the proposed traffic signal controller can make real-time adaptation in the sense that it responds effectively to the changing road dynamics. These road dynamics are simulated by the Green Light District (GLD) vehicle traffic simulator that is the testbed of our traffic signal control. We have implemented the Intelligent Driver Model (IDM) acceleration model in the GLD traffic simulator. The change in road conditions is modeled by varying the traffic demand probability distribution and adapting the IDM parameters to the adverse weather conditions. Under the congested and free traffic situations, the proposed multi-objective controller significantly outperforms the underlying single objective controller which only minimizes the trip waiting time (i.e., the total waiting time in the whole vehicle trip rather than at a specific junction). For instance, the average trip and waiting times are lower  8 and 6 times respectively when using the multi-objective controller.
Keywords:
adaptive optimization, multi-objective optimization, reinforcement learning, exploration, traffic signal control, cooperative multi-agent system

In recent years, bus priority techniques for signals controlled by traffic management centers have become a viable alternative to reduce passenger delays at signalized intersections, especially in mixed traffic corridors. However, before any bus signal priority strategy is deployed in such corridors, the impacts on the different users of the system should be evaluated. The main objective of this work was to assess the operational performance of passive and active bus priority techniques in fixed and real-time signal systems of one of the main arterial corridors in Fortaleza, Brazil. As a secondary objective, it also evaluated the operational benefits of a SCOOT adaptive signal control system, comparing it with well-adjusted fixed-time plans optimized by TRANSYT, for periods of medium and high traffic volumes. In the evaluation of alternative scenarios, the
following performance measures were considered: vehicle delay and number of stops simulated by SCOOT, as well as bus and automobile travel times observed in the field during the operation of each scenario. The results did not favor the adoption of passive and active priority schemes in the studied corridor; this led to the conclusion that SCOOT’s real-time control, programmed for a good signal progression of the general traffic (buses and automobiles), is the best signal control strategy for an arterial corridor similar to the one under analysis.

To obtain several parameters of signalized intersection, such as green time, width of Left Turn on Red (LTOR) lane, and number of lane, which comply with the specific value of degree of saturation, the manually calculation procedure with or without software Indonesian Highway Capacity Manual (IHCM) must be iterative several times. By using Microsoft Excel Solver, the process of calculation to obtain several parameters of signalized intersection which comply with the specific value of degree of saturation become easier and faster. This simple application in Microsoft Excel using common function and Solver add-in can also modify to comply several others specific performance measures, such as vehicle operation cost, time value, etc. Abstrak Dalam perhitungan kinerja simpang bersinyal, untuk mendapatkan alokasi waktu hijau, lebar lajur belok kiri langsung (Left Turn on Red, LTOR), dan jumlah lajur, yang memenuhi syarat nilai derajat kejenuhan tertentu, perhitungan dengan maupun tanpa menggunakan software Kapasitas Jalan Indonesia (KAJI) harus diulang dari awal. Tujuan penelitian ini adalah membuat aplikasi spreadsheet untuk mempermudah dan mempercepat memperoleh nilai batasan kondisi geometrik jalan serta alokasi waktu hijau pada suatu simpang bersinyal yang memenuhi batas nilai DS tertentu. Aplikasi spreadsheet yang dibuat dengan memanfaatkan fasilitas add-in Solver pada Microsoft Excel juga dapat dikembangkan lebih lanjut dengan menambahkan berbagai batasan optimasi kinerja, antara lain, biaya operasional kendaraan (BOK), nilai waktu, dan lain sebagainya. Kata Kunci: simpang bersinyal, optimasi, solver PENDAHULUAN Kinerja sistem jaringan jalan tidak saja dipengaruhi oleh kinerja ruas jalannya, tetapi juga oleh kinerja setiap persimpangannya. Meskipun kinerja ruas jalan dari suatu sistem jaringan jalan sudah baik, jika kinerja persimpangannya sangat rendah maka kinerja seluruh sistem jaringan jalan tersebut akan menjadi rendah pula (Tamin, 2000). Kinerja ruas jalan dan persimpangan jalan dipengaruhi oleh besarnya arus lalu lintas. Selain itu, kinerja ruas jalan dipengaruhi oleh lebar jalan, pembagian arah, hambatan samping dan ukuran kota. Sedangkan kinerja persimpangan jalan dipengaruhi oleh adanya pergerakan konflik (karena ada kendaraan yang belok kanan), nilai arus jenuh, waktu hijau efektif dan waktu siklus. Berdasarkan Manual Kapasitas Jalan Indonesia/MKJI (1997), analisis persimpangan dibagi menjadi tiga, yaitu persimpangan tak bersinyal, bagian jalinan dan persimpangan bersinyal.

This paper presents a method of addressing stochastic variation at closely spaced signalized intersections to provide secondary coordination to “minor” movements with significant traffic volumes. A neuro fuzzy signal control system was designed in this study to manage a non-coordinated movement to avoid queue spillback. Building on the conventional actuated-coordinated control system, the neuro-fuzzy controller does not lose the benefit of the primary coordination of the conventional controller but establishes a “secondary coordination” between the upstream coordinated phase (through phase) and the downstream non-coordinated phase (left-turn phase) on the basis of areal-time traffic demand. Under the neuro-fuzzy signal control, the traffic from the upstream intersection can arrive and join the queue at the downstream left-turn lane and be served in a timely fashion and thus reduce the likelihood of being delayed at the downstream intersection. The simulation results indicate that the neuro-fuzzy signal control consistently outperformed the conventional actuated-coordinated controller in terms of reduction in systemwide average delay and number of stops per vehicle under a wide range of traffic volumes by nearly 20% under heavier demand conditions.

Most isolated signalised intersections in the suburban area are controlled using a vehicle-actuated system due to its ability to allocate green times based on actual demand. The system use detectors that are located in travel lanes. These detectors sense the presence or passage of vehicles (Cassidy, et al. 1996). This information is used by the controller to extend the existing green time as shown in Figure 1.

The equations developed by Webster in his famous 1958 report[10] are still the basis of traffic signal planning today. They are being used in handbooks like the HCM and similar instruments world-wide. However, the handbook approach typically works with approximations to the original equations which have stood the test of time, but may nevertheless not be the best to be done today. This work analyzes Webster’s approach and advocates a more modern use of it which utilizes the tremendous advances in computer hardware and software. This is being done by comparing approximations to exact solutions, and by a comparison between various models and Webster’s equations itself. It is shown that there can be significant differences in the calculation of optimal cycle times and consequent delay times.

Traffic congestion has now become a very critical issue particularly in metropolitan Dhaka, the capital of Bangladesh. Because of the rapid socio-economic changes and increase in population, the city expanded dynamically without any planning and control. Such rapid and uncontrolled developments have created an unacceptable degree of disparity in the transportation demand and supply scenario, which resulted in traffic jam. In order to alleviate traffic congestion, it is needed to improve traffic operating condition of the existing road network by applying traffic management tools. In this research, an attempt has been interpreted to ameliorate the traffic operating condition of Mirpur road (Manik Mia Avenue, Russell Square and Rapa Plaza Intersection) by setting appropriate signal timing at the intersections along this road. CORSIM simulation software is used to perform the analyses. Collection of traffic volume data is conducted by video recording. To obtain maximum flow survey is done during the peak periods. The existing condition of the Mirpur road (Russell Square to Aarong) is modeled in CORSIM and calibrated with the field data. It is found that along the road Russell Square, Rapa Plaza and Aarong intersections are overly congested. Optimum signal timings are set at the intersections to make the traffic operating condition better. Satisfactory results has been obtained as the average speed of the vehicles increased and delay time decreased for almost every link as well as for the whole road network. Bus Rapid Transit (BRT) and Metro Rail Transit (MRT) system are going to be implemented in Dhaka city to reduce the traffic jam. Along with this road way capacity improvement should be the primary concern to abate the intensity of traffic in roads.

This paper focuses on real-time traffic management facilitated by modern telecommunication technologies and advanced real-time optimization algorithms. The discussion begins with a recent European project that provides a real-time decision support system for the reduction of traffic congestion and emissions. The work ow and techniques involved therein are explained and issues and potential gaps identified. A more generic real-time decision-making framework based on decision rules and distributionally robust optimization is then introduced. The paper illustrates the wide applicability and unique advantages of such a framework with case studies on responsive signal control, use of an adaptive variable message sign, and air traffic management.

Several factors are usually considered when recommending left-turn phasing at signalized intersections: the volume of through traffic opposing the left-turn maneuvers, the volume of left-turn traffic, the speed of the opposing traffic, sight distance, crash history, and cycle length. The goal of this paper is to find the optimal signal phasing designs that can improve efficiency and safety of the particular signal design in the field in order to test current recommendations and guidelines on left-turn treatments. The researchers assess the impact of the number of rings in a phasing structure, the number of phases and whether to adopt lead or lag phasing on the safety and efficiency of a four-legged signalized intersection. A comprehensive set of 338 signal phasing designs is developed and the impacts are modeled using VISSIM microsimulation. The efficiency of signal phasing sets is evaluated in VISSIM, while the safety aspect is assessed using SSAM software. Optimal scenarios are tested using sensitivity analysis for various traffic flows. The findings show that the current left-turn guidelines may over-protect left-turners which results in reduced efficiency and left-turn delays that can be increased by up to 25%. Revised guidelines should contribute to the decreased delays for left turning vehicles, while keeping the safety standards high.

Different left-turn signal phasing designs have been shown to result in significant differences in traffic delay, intersection capacity, and safety level. Past studies on this matter focused mainly on phasing design at isolated intersections. This paper advanced the theory by investigating the traffic flow patterns at two coordinated signalized intersections, comparing the differences in delay based on leading and lagging left-turn signal phase designs, identifying the main control factors, and establishing simple criterion for choosing between the leading and lagging left-turn sequences at coordinated signal intersections.

It was found that for coordinated signalized intersections, leading or lagging designs do not differ in terms of through traffic delays. Instead, their strength/weakness resides mainly in left-turning traffic delay. Lagging design for the downstream signal generates less delay than leading design no matter which design was used for the upstream signal. Lagging design at both intersections yields best results in terms of overall intersection delay.

The design of signal timing along multiple intersections of arterial Eagle Road in Meridian Idaho was accomplished in TRANSYT-7F 11.3 and Vistro 2.0.  TRANSYT-7F and Vistro are traffic simulation and signal timing optimization programs.  The results were compared for six intersections: Chinden Boulevard, Bristol Heights Road, McMillan Road, Wainwright Drive, Ustick Road, and Fairview Avenue along the Eagle Road corridor.  Performance measures used to compare the results from each program included Level of Service (LOS), control delay, and queue lengths.  Based on the performance measures of LOS, control delay, and queue lengths, TRANSYT-7F outperformed Vistro.  Using the results of TRANSYT-7F, recommendations for corridor improvements were made to increase performance measures. The network was recoded and optimized with the improvements.

In this article, we propose a new analytical traffic flow model for traffic dynamics at signalized intersections. During each cycle, both the arrival and the departure traffic are approximated by three distinct traffic streams with uniform density. Because of the similar representation of the arrival and the departure traffic, the results from a single intersection can easily be extended to a series of intersections. The number of parameters of the model is tractable, leading to analytical solutions of the problem. We prove that the total delay of one-way traffic is a quasi-convex function in the offset between consecutive traffic cycles and derive analytically the optimal signal control corresponding to different profiles of arrival densities. This allows timely adjustments of the control as congestion evolves throughout the day. We also study how different density profiles evolve in a corridor, from one intersection to the downstream one, if there is no traffic from/to side streets. We find that all density profiles eventually lead to one profile after a few intersections. This corresponds to a green wave, in agreement with physical intuition. Finally, we test the model against data from microsimulation using CORSIM. Vehicle delay predicted by the model is shown to be close to that from the microsimulation.

In most traffic signal optimization problems, pedestrian traffic at an intersection receives minor consideration compared to vehicular traffic, and usually in the form of simplis-tic and exogenous constraints (e.g., minimum green time). This could render the resulting signal timings sub-optimal especially in dense urban areas with significant pedestrian traffic, or when two-stage pedestrian crosswalks are present. This paper proposes a convex (quadratic) programming approach to optimize traffic signal timings for an isolated intersection with one-and two-stage crosswalks, assuming undersaturated vehicular traffic condition. Both vehicle and pedestrian traffic are integrated into a unified framework, where the total weighted delay of pedestrians and vehicles at different types of crosswalks (i.e. one-or two-stage) is adopted as the objective function, and temporal and spatial constraints (e.g. signal phasing plan and spatial capacity of the refuge island) are explicitly formulated. A case study demonstrates the impacts of incorporating pedestrian delay as well as geometric and spatial constraints (e.g., available space on the refuge island) in the signal optimization. A further analysis shows that a two-stage crosswalk may outperform a one-stage crosswalk in terms of both vehicle and pedestrian delays in some circumstances.

The paper presents results of our research on application of genetic algorithms to the problem of finding good configurations of traffic signals offsets in a road network (Traffic Signal Setting problem). We tested algorithms on a large road network-realistic map of Warsaw acquired from the OpenStreetMap project. The main research tool was the software Traffic Simulation Framework developed by the first author. To speed up experiments we employed a high-performance computing cluster at the University of Rzeszów-sessions of experiments were supervised by the second author.

Highway capacity is a scarce resource which needs to be allocated economically. Traffic control at signalized intersections involves balancing competing demands of different traffic streams for limited capacity at the intersection. Similarly, progression schemes on two-way arterial streets involve a compromise between the demands of opposing and competing traffic streams along the arterial. In these and similar cases there is a tradeoff between the performance potential that each traffic stream can attain and there is a need for determining the most effective signal controls under given circumstances. This phenomenon is akin to tradeoffs in production capabilities of economic systems that gives rise to the well-known Production Potential Frontier. We develop similar paradigms for signalized intersections and for coordinated arterial streets which are called Performance Potential Frontiers (PPF). We introduce two types of PPFs: the Transition Potential Frontier (TPF) for a single intersection and the Progression Potential Frontier (PPF) for a coordinated arterial street. The TPF measures the favorability of vehicles traversing an intersection as a function of the right-of-way allocation to competing demands. Similarly, the PPF quantifies the advantages in performance that can accrue to competing traffic flow streams along an arterial. Three approaches are introduced to assess the performance of arterial streets using the HCM Level-of-Service (LOS) concept. These approaches incorporate systematic accounting of the effects of coordination through the Cyclic Coordination Function (CCF) and the corresponding Coordination Adjustment Factor (CAF).

This paper describes a new computer method and program, MITROP (Mixed Integer TRaffic OPtimisation), which is designed to simultaneously optimize all the traffic signal control variables in a road network, including the cycle time, the green times and the offsets. The optimization objective is to minimise the disutility for travel in the network and the technique used is mixed-integer linear programming. The main features of the method are described.

A methodology for obtaining a signal  plan ( cycle and green times ) is formulated for a Isolated traffic signal , The plan can cover from a few hours to several days. This methodology explicitly considers the variations of traffic flows that occur during the period of analysis for each of the approaches to the intersection .

Traffic congestion is a challenging problem faced in everyday life. It has multiple negative effects on average speed, overall total travel time, and fuel consumption; in addition, it is a primary cause of accidents and air pollution. Hence, comes the need for an intelligent reliable traffic control system. The objective of this paper is to optimize the overall traffic congestion in freeways via multiple ramps metering controls. An optimal freeway operation can be reached if we always keep the freeway density within a small margin of the critical ratio for maximum traffic flow. In this paper, we propose a multi-agent reinforcement learning control system for ramp metering. Our system introduces a new microscopic framework at the network level based on collaborative Markov Decision Process modeling and an associated cooperative Q-learning algorithm. The technique incorporates payoff propagation (max-plus algorithm) under the coordination graph framework, particularly suited for optimal control purposes. The proposed system provides three control designs: fully independent, fully distributed, and centralized; suited for different network architectures. Our framework was extensively tested in order to assess the proposed model of the joint payoff, as well as the global payoff. We conducted experiments with heavy traffic flow under the renowned VISSIM traffic simulator so as to evaluate the proposed framework. The experimental results show that we achieved a significant decrease in the total travel time and an increase in the average speed -when compared with the base case- while maintaining an optimal traffic flow.
Keywords: Multi-agent system, intelligent transportation system, adaptive traffic control, freeway, ramp metering, max-plus.

We propose a two-stage, on-line signal control strategy for dynamic networks using a linear decision rule (LDR) approach and a distributionally robust optimization (DRO) tech- nique. The first (off-line) stage formulates a LDR that maps real-time traffic data to optimal signal control policies. A DRO problem is solved to optimize the on-line performance of the LDR in the presence of uncertainties associated with the observed traffic states and ambiguity in their underlying distribution functions. We employ a data-driven calibration of the uncertainty set, which takes into account historical traffic data. The second (on-line) stage implements a very efficient linear decision rule whose performance is guaranteed by the off-line computation. We test the proposed signal control procedure in a simulation environment that is informed by actual traffic data obtained in Glasgow, and demonstrate its full potential in on-line operation and deployability on realistic networks, as well as its effectiveness in improving traffic.

In this paper, we take a self-scheduling approach to solving the traffic signal control problem, where each intersection is controlled by a self-interested agent operating with a limited (fixed horizon) view of incoming traffic. Central to the approach is a representation that aggregates incoming vehicles into critical clusters, based on the non-uniformly distributed nature of road traffic flows. Starting from a recently developed signal timing strategy based on clearing anticipated queues, we propose extended real-time decision ...