Multimodeling: A Preliminary Case Study
نویسندگان
چکیده
We take a pre-existing Statecharts model of a simple traffic light controller and re-implement it in Ptolemy II. This exercise reveals that Statecharts can be usefully conceptualized as a hierarchical combination of two distinct models of computation (MoCs), finite state machines (FSMs) and synchronous/reactive (SR). Once conceptualized this way, we can add additional MoCs to the mix. We illustrate this by adding a discrete-event (DE) model of the environment in which the traffic light operates. We then construct a second model of a deployment of the system on two microcontrollers communicating wirelessly, showing that we can effectively leverage both DE and an extension in Ptolemy II that supports modeling of wireless communication networks. This exercise reveals that even though the original model was intended to be a purely functional model, it in fact imposes constraints on the implementation. The model had to be refactored to get a distributed deployment model. Finally, we show that the portions of the models defining the control logic of the lights can be shared between the functional and deployment models using actor-oriented classes. This eases maintenance of the models.
منابع مشابه
Model Engineering using Multimodeling
We study the simultaneous use of multiple modeling techniques in the design of embedded systems. We begin with a pre-existing Statecharts model of a simple case study, a traffic light for a pedestrian crossing, using it to illustrate the need for multimodeling and the pitfalls. The original model combines two distinct models of computation (MoCs), finite state machines (FSMs) and synchronous/re...
متن کاملStochastic H∞ Control Problem with State-Dependent Noise for Multimodeling Systems
Abstract—In this paper, stochastic H∞ state feedback control with state-dependent noise for multimodeling systems is addressed. After establishing the asymptotic structure of the stochastic multi-modeling algebraic Riccati equation (SMARE), an iterative algorithm that is based on Newton’s method is established. A high-order state feedback controller by means of the obtained iterative solution i...
متن کاملOn Design of a Coupling Component for Parallel Multimodeling
A component-based simulation framework is a favorable choice for many multi-scale simulations, which requires dedicated coupling to support flexible data communication/exchange between components, computational parallelism and on-demand dynamic load balancing. In this paper, authors first present the considerations in the course of design and implementation of a coupling component to support in...
متن کاملMultimodeling Representation for Physical Systems Application to PWR Primary Coolant Loop
This paper deals with the multimodeling principle to represent physical systems. It shows that when we apply this approach to a Pressurized Water Reactor (PWR) nuclear power plant, we face some difficulties especially in representing behavioral model. In order to solve this problem we propose a solution consisting in introducing bond graph concept at the behavioral level. Model-based reasoning ...
متن کاملEvaluating Conceptual Coherence in Multi-Modeling Techniques
Meta-modeling is a well-known approach for capturing modeling methods and techniques. A meta-model can serve as a basis for quantitative evaluation of methods and techniques. By means of a number of formal metrics based on the meta-model, a quantitative evaluation of methods and techniques becomes possible. Existing meta-modeling languages and measurement schemes do not allow the explicit model...
متن کامل