Jarkko Niittymäki

Signal control using fuzzy logic

Applications with fuzzy logic being used in controlling traffic signals have been designed already since the 1970s. The strength of fuzzy logic lies in its capability of simulating the decision-making process of a human, a process that is often difficult to define with traditional mathematical methods. The results of FUSICO-project have indicated that the fuzzy traffic signal control can be the potential control method for signalized intersections.

Application of Fuzzy Logic to the Control of a Pedestrian Crossing Signal

Fuzzy logic is known to be suited for dealing with a complex optimization problem with many objectives, many constraints, unclear input information, and vague decision criteria. Controlling the timing of a traffic signal falls in this category of problem. Fuzzy logic is introduced for controlling the timing of a pedestrian crossing signal. The controller is designed to emulate the decision process of an experienced crossing guard. The performance of this control is tested against two types of conventional demand-actuated control: one that uses the traditional green extension and the other that uses modified extension rules. The criteria for evaluation are delays to the pedestrians and the vehicles, and the percentage of vehicles that are stopped. The fuzzy logic controller finds a compromise between two conflicting objectives: minimization of pedestrian delay and minimization of vehicular delay and stops. The evaluation was performed using a microscopic simulation called HUTSIM developed at the Helsinki University of Technology. The fuzzy logic controller performs equally well as or better than conventional demand-actuated control without requiring many parameter settings. Furthermore, the control rules are simple and a compilation of rational decision processes is expressed in natural language.

Traffic signal control on similarity logic reasoning

The main intention in this study is to tie fuzzy reasoning to many-valued logic framework; a Lukasiewicz many-valued logic similarity based fuzzy control algorithm is introduced, and tested in three realistic traffic signal control systems. The results are compared to fuzzy control systems where the inference is based on standard Matlab Fuzzy Logic tooboxs Mamdani-style system. The compared traffic signal control modes are signalized pedestrian crossing and multi-phase signal control with phase selection. The statistical significance in differences of expectations obtained by different control schemes has been tested by two-sided approximate Students test on significance level α = 0.01. Simulations made by HUTSIM traffic signal simulator show that the performances of the new control algorithm and that of standard Mamdani-style algorithm are almost equal. However, if traffic density is high then the new algorithm gives significantly better statistical results. Moreover, a stronger mathematical approach offers a natural smoothness test.

Installation and experiences of field testing a fuzzy signal controller

Abstract This paper describes the installation of a fuzzy signal controller (FSC) at a real intersection. The results of a vehicle-actuated control system with a fuzzy-control system using microscopic simulation and field measurements have been compared. The results indicate that the fuzzy control is very competitive against traditional vehicle-actuated control, if traffic volumes are higher than low-demand. The benefit of fuzzy logic lies in its ability to handle linguistic information by representing it as a fuzzy set. The simple algorithm structure, the savings of material costs and the low installation and maintenance costs are important advantages. The results of this paper prove that the FSC can be installed in real infrastructure and that fuzzy algorithms can be more effective than traditional vehicle-actuated control.

General fuzzy rule base for isolated traffic signal control-rule formulation

Traffic signal control is one of the oldest applications of fuzzy logic, at least in transportation engineering. The aim of this paper is to present a systematic approach to fuzzy traffic signal control and to derive the linguistic control rules based on expert knowledge. Traffic signal programming is generally divided into two problems: firstly, the choice and sequencing of signal stages to be used, and secondly, optimizing the relative lengths of these stages. The rule bases for both problems are introduced in our paper. The results of tested rule bases and field tests of fuzzy control have been promising. The fuzzy signal control algorithms offer better measures of effectiveness than the traditional vehicle‐actuated control.

Saturation Flows at Signal-Group-Controlled Traffic Signals

The main goal of this research was to update the basic saturation flow values of signalized intersections. The secondary goal was to analyze the effects of certain external factors (such as weather, road, and traffic conditions) on saturation flow. The updating is based on extensive field measurements and simulations. Altogether, about 39,000 queues were observed in this study. Field measurements at 30 locations were made according to the method described in the Highway Capacity Manual and simulations were done with the Helsinki University of Technology HUT-SIM simulator, which was calibrated and carefully validated for Finnish road conditions. A summary of calibration parameters is also presented. The new base value for straight-through lanes is 1, 940 vehicles per hour; the previous value was 1, 700 vehicles per hour. In general, the updated saturation flow values of different lane types are 5 to 20 percent larger than the previous base values. The saturation flow models of different lane types are described. The effects of geometric and traffic composition factors, such as percentage of turning vehicles, traffic composition, lane width, and approach grade, were examined and modeled. Effects of weather, road surface, light conditions, and speed level were also analyzed. The drop in saturation flow was about 20 to 30 percent under slippery road and snowy conditions. In rainy conditions, the drop was smaller, about 10 percent. The effect of speed on saturation flow is also described. The most important results of this 2-year project are the saturation flow values for different lane types, knowledge of the effect of external factors (especially during winter), and the large database, which can be used for other purposes. The possibility of using special signal control programs under bad road conditions is discussed. With these kinds of programs, better safety and higher capacity can be achieved.

Fuzzy logic two-phase traffic signal control for coordinated one-way streets

The specific goal of this study was to compare the differences in various traffic signal control algorithms when simulating two consequent one-way intersections with no turning traffic. The opportunities offered by fuzzy logic for controlling traffic are the subject of our FUSICO (FUzzy logic SIgnal group COntrol) research project. The compared algorithms used in the simulations were coordinated fixed-time signal group control, traffic-actuated gap-seeking non-coordinated signal group control, traffic-actuated FUSICO, and a combination of standard traffic-actuated control and a FUSICO controller. This study was our first attempt to use fuzzy methods in coordinated traffic signals; the results were promising. All previous results of the FUSICO project have indicated better overall efficiency than traditional vehicle-actuated control.

Traffic signal controller based on fuzzy logic

The main goal of this study is to introduce a prototype of a new traffic signal controller based on fuzzy logic. Technical information as well as the basic outline of the software is introduced. The fuzzy inference part of the controller is described in details and used fuzzy methods are introduced in briefly. In the second part of the study, the results of the before-after measurements of the field test are introduced.

New methods for traffic signal control-development of fuzzy controller

The main goal of the study is to introduce two fuzzy control systems for traffic signal control. Two different fuzzy control methods are tested in both control systems and their mutual goodness is compared. The tested fuzzy inference methods are a traditional Mamdani-type system and a fuzzy inference method based on maximal fuzzy similarity. For the fuzzy inference method based on maximal fuzzy similarity there is introduced an extension for which fuzzy inference systems where both input and output spaces are fuzzy can be constructed. The second goal of this study is to introduce a method for stability analysis of fuzzy control systems, fuzzy rulebase and membership functions.

Fuzzy traffic signal control in major arterials

The opportunities offered by fuzzy logic for controlling traffic are the subject of our FUSICO-research project in Finland at the Helsinki University of Technology. In the previous studies of FUSICO-project, traffic signal control based on fuzzy logic has successfully been tested in urban intersections. The promising results of simulations and field installation suggest that the fuzzy control method would be competitive also in high-class arterial roads with high safety and efficiency demands. Encouraged by this, we are developing a fuzzy control method ITCARI for arterial intersections. The pilot version of ITCARI-control is installed at an intersection in the city of Tampere, and the field measurements will be done during summer 2001. First simulation studies of the Tampere case indicate competitive performance of the fuzzy control method on arterial roads.