Michel dos Santos Soares

Real-time scheduling of batch systems using Petri nets and linear logic

This paper presents an approach to model, design and verify scenarios of real-time systems used in the scheduling and global coordination of batch systems. The initial requirements of a system specified with sequence diagrams are translated into a single p-time Petri net model representing the global behavior of the system. For the Petri net fragments involved in conflicts, symbolic production and consumption dates assigned to tokens are calculated based on the sequent calculus of linear logic. These dates are then used for off-line conflict resolution within a token player algorithm used for scenario verification of real-time specifications and which can be seen as a simulation tool for UML interaction diagrams.

Responsive traffic signals designed with petri nets

Traffic-responsive techniques make use of real-time measurements acquired with sensors to calculate in real-time suitable settings. Traffic signal systems that react to changing traffic conditions are an important component for improving transportation efficiency. The dynamic behavior of a group of traffic signals controlling a network of intersections is a complex discrete event system that can be modeled using Petri nets. In this paper, one purpose is to design a mechanism based on Petri nets with time interval associated to places to extend the green time to a main road depending on the demand of non-priority roads. Another objective is to try to allow green phases for a sequence of intersections in a small network in order to improve traffic flow for a platoon of vehicles. Modular characteristics of Petri nets are used to address complexity and design models in increasing levels of detail.

Human factors in system reliability: lessons learnt from the Maeslant storm surge barrier in The Netherlands

The Maeslant storm surge barrier in the Netherlands is an interesting case in system reliability, first because of the great effort that has been put into making its operation reliable and into assessing its reliability, and second, because it has characteristics that make reliability assessment extremely hard. From its history a number of interesting conclusions can be drawn, of which the most important one is that there is no straightforward, definitive solution to reliability, but reliability is obtained and maintained in a continuous process of improvement. Other conclusions are that humans cannot be excluded from the operation or decision-making in systems such as the Maeslant barrier, that all methods for improving system reliability are most effective when the people involved are sharply aware of each methods limitations and that a continuous, open process of consulting a variety of experts is crucial to obtain the best possible reliability.

A distributed simulator for road network control

This paper is about a distributed implementation of a road traffic simulator. The simulator will be used in an integrated, operational system for road traffic monitoring, visualization, prediction and control. The distributed implementation is needed in order to make the simulator scalable (especially with respect to road network size) and to make it comply with real-time requirements stemming from on-line use in control and decision support systems, with emphasis on network control rather than local control. The underlying traffic model is a first order, multi-class model. The implementation has been tested with the network of the Amsterdam A10 beltway.

A hierarchical model and implementation architecture for road traffic control

Control of traffic in a network of roads may involve a large number of individual control loops, with various scopes and time scales, some of them working locally, such as traffic signals at a crossing, some of them coordinating a number of local control loops within its scope. In order to organize such a set of control loops, this paper proposes a hierarchical network model which is primarily based on a modularity property of networks. One of the prime applications of this hierarchical model is to derive an implementation architecture for the development of operational control systems, including the necessary hardware and software infrastructure, in order to achieve highly flexible control systems and thereby support the research in road traffic network control. The primary goal of this paper is to list the research issues of this approach.

Intelligent network traffic control by integrating top-down and bottom-up control

In the Netherlands, the common approach for road network traffic control is top-down control by applying so-called control schemes. However, it has several important drawbacks due to its centralized control nature. Only traffic control centers will decide the control scheme according to patterns observed in traffic. As the system is distributed over a network, bottom-up control is a good complement to top-down control by applying multi-agent distributed control. Transforming these elements into agents will make traffic respond much faster to change. This paper presents an approach to integrating both the bottom-up and the top-down control for road network traffic control systems. The combined approach can have all the advantages of both approaches and avoid their disadvantages. This approach has been applied in the operational traffic control system for the Dutch city of Alkmaar. It was developed at the Dutch traffic management company Trinité Automatisering B.V.

Architecture-based development of road traffic management systems

Road traffic problems, such as congestion and accidents, are high in many countries, leading to economic losses, environmental damage due to increased pollution, waste of time in congestion and lives lost. In order to reduce these problems, Road Traffic Management Systems (RTMS) are applied, for example, in activities such as controlling, monitoring, and visualizing traffic in motorways and urban roads. The development of RTMS is challenging. RTMS are large socio-technical systems, which means that organizational and regulatory policies, rules, processes and constraints have to be taken into consideration. This framework needs to be represented in what is called the domain architecture. However, defining the domain architecture relates more to the RTMS domain than to the technical solution. The software architecture is fundamental for the organization of the system, comprising the sub-systems, important software components and the relationship between them and the environment, and to document all important technical decisions. In this article, both the domain and the software architecture are proposed for developing RTMS. The approach is used in practice as shown by the HARS case study.

A hierarchical network model for road traffic control

Control of traffic in a network of roads may involve a large number of individual control loops, with various scopes and time scales, some of them working locally, such as traffic signals at a crossing, some of them coordinating a number of local control loops within its scope. In order to organize such a set of control loops, this paper proposes a hierarchical network model which is primarily based on a modularity property of networks. One of the prime applications of this hierarchical model is to derive an implementation architecture for the development of operational control systems, in order to achieve highly flexible control systems and thereby support the research in road traffic network control. The primary goal of this paper is to list the research issues in this approach.

Modeling and analysis of discrete event systems using a Petri net component

The dynamic behavior of a group of traffic signals controlling a network of intersections is a complex discrete event system that can be modeled by Petri nets. The approach used in this article proposes a components-based design, which increases modularity, address complexity and is a good practice according to modern Software Engineering. All the main system elements are specified based on the proposed Petri net component with time interval associated to places. The specified models are simulated through the common token player algorithm. Formal analysis using place invariants and theorem proving are applied to verify models soundness and reason on specific scenarios.

Application of publish-subscribe middleware to road traffic measurements visualization

Visualization of the effects of control measures is important for decision support in complex networks. The example in this article is a visualization system for road traffic measurements, such as intensity, density, waiting time and velocity, at junctions in networks controlled by traffic signals. The system was developed based on a layered implementation architecture, which has the advantage of abstracting lowlevel details during development, and using an object-oriented approach. In order to cope with the complexity of the system due to its distributed nature and real-time constraints, we decided to apply publish-subscribe middleware for components communication, which offers some important advantages, such as scalability, flexibility, and improved performance.