Fernando J. Barros

Modeling formalisms for dynamic structure systems

We present a new concept for a system network to represent systems that are able to undergo structural change. Change in structure is defined in general terms, and includes the addition and deletion of systems and the modification of the relations among components. The structure of a system network is stored in the network executive. Any change in structure-related information is mapped into modifications in the network structure. Based on these concepts, we derive three new system specifications that provide a shorthand notation to specify classes of dynamic structure systems. These new formalisms are: dynamic structure discrete time system, dynamic structure differential equation specified systems, and dynamic structure discrete event system specification. We demonstrate that these formalisms are closed under coupling, making hierarchical model construction possible. formalisms are described using set theoretic notation and general systems theory concepts.

Dynamic structure discrete event system specification: a new formalism for dynamic structure modeling and simulation

Traditional simulation methodologies do not support changes in model structure during a simulation run. Current methodologies support only changes in model descriptive variables. Changes in structure are thus forced to be represented at the simple behavioral level. Many models are better represented at both behavioral and structural level. We present a new simulation formalism that fully supports changes in model structure and its closure under coupling. The Dynamic Structure Discrete Event System Specification formalism (DSDEVS) supports a new simulation paradigm, structural simulation, as opposed to conventional trajectory simulation. This new formalism supports changes in structure to the full extent, ranging from simple model/connection add/deletion to the exchange of models between network of models. The DELTA simulation environment, an implementation of the DSDEVS formalism, is briefly described.

Abstract simulators for the DSDE formalism

We present the DSDE (Dynamic Structure of Discrete Events) formalism, a methodology for representing discrete event systems that change structure dynamically. We prove that the DSDE formalism is closed under coupling and that it can be used to construct hierarchical and modular models. The abstract simulators which are necessary to execute dynamic structure models are also presented. These simulators allow a description of models independent of the actual simulation procedure, and thus encourage model reuse.

Dynamic structure multiparadigm modeling and simulation

This article presents the Heterogeneous Flow System Specification (HFSS), a formalism aimed to represent hierarchical and modular hybrid flow systems with dynamic structure. The concept of hybrid flow systems provides a generalization of the conventional concept of hybrid system and it can represent a whole plethora of systems, namely: discrete event systems, multicomponent and multirate numerical methods, multirate and multicomponent sampling systems, event locators and time-varying systems. The ability to join all these types of models makes HFSS an excellent framework for merging components built in different paradigms. We present several examples of model definition in the HFSS formalism and we also exploit the ability of the HFSS formalism to represent mutirate numerical integrators.

Towards a theory of continuous flow models

We present the continuous flow system specification formalism (CFSS). This formalism provides a digital computer approach to the representation of continuous systems based on a broad interpretation of the sampling concept. Continuous flow models permit to represent systems with both time-varying and multirate sampling. The equivalent dynamic system of a CFSS model is defined. Based on the CFSS formalism a multirate integration method is developed.

Multimodels and dynamic structure models: an integration of DSDE/DEVS and OOPM

Constructing models of systems that change their structure over time has proved to be a challenging problem, with several proposed solutions. We present two of these approaches and discuss their integration. The parallel Dynamic Structure Discrete Event (DSDE) system specification is based on systems-theoretic concepts and provides a formal specification of variable structure models. Object-oriented physical modeling (OOPM) extends the design approaches of software engineering and employs both static and dynamic models to describe physical objects. OOPM is based on the multi-model concept to support adaptive structures. The integration is achieved by representing multi-models within the dynamic-structure Discrete Event System Specification (DEVS) formalism. Since the latter has been implemented in HLA (High-Level Architecture) compliant form, the integration has the practical consequence that multi-models can be directly mapped into distributed simulations.

Modeling and Simulation of Dynamic Structure Heterogeneous Flow Systems

The author presents the Heterogeneous Flow System Specification (HFSS), a formalism suited to represent combined (discrete/continuous) systems. The new formalism is based on the concept of sampling that is described in general terms to include multirate sampling systems. The HFSS formalism permits the construction of continuous/discrete input/output models in a modular and hierarchical manner. Hierarchical systems have a dynamic structure nature that can be changed over time. To show an application of the HFSS formalism, the author presents a new formalism to numerically solve differential equations with discontinuities using multirate integration methods. Applications of combined and dynamic structure systems are presented.

Forest fire modelling and simulation in the DELTA environment

Abstract Traditional simulation methodology supports only changes in models state variables. Some models are better expressed by a combination of both changes in state variables and changes in structure. Dynamic Structure Discrete Event Specification (DSDEVS) is a recently introduced modelling and simulation formalism that provides full support for representing models with time varying structure. The DELTA simulation environment is an implementation of the DSDEVS formalism and provides full support to Structural Simulation. We show the advantages of the Dynamic Structure Cellular Automata describing the model of forest fire spreading and its implementation in the DELTA environment.

Increasing Software Quality through Design Reuse

The development of new applications based on existing design and code increases software quality, since tested assets are more likely to exhibit fewer errors than software developed from scratch. Modular and hierarchical software topologies promote reuse by allowing the development of applications based on independent software units that can be arbitrarily interconnected. In this paper we introduce a different dimension to software reuse based on topology. This kind of reuse is termed here by inheritance of topology and permits the use of existing software topologies to define new ones without incurring in the cost of creating designs from scratch. To illustrate this concept we employ Connectons, a software topology that combines a modular and hierarchical software definition with the request reply paradigm. In this paper we show that inheritance of topology promotes software quality and in particular we show that it provides a better support for reuse than the creational design patterns Factory Method and Abstract Factory.

Variable DEVS-variable structure modeling formalism: an adaptive computer architecture application

Conventional modeling theory gives support only for representing model behavior, providing little aid for describing changes in model structure. Some models are better represented by changes in their structure. Instead of forcing this changes to be represented at the simple behavioral level, a strong theoretical support is needed to allow the representation of structural changes in a natural way. We present a modeling methodology for representing variable structure systems. Examples of such systems include adaptive computer architectures, ecological systems, fault tolerating computers. We describe an application of this methodology to the modeling and simulation of an adaptive computer architecture.<<ETX>>