%0 Journal Article
%T Model-Driven Development of Automation and Control Applications: Modeling and Simulation of Control Sequences
%A Timo Veps£żl£żinen
%A Seppo Kuikka
%J Advances in Software Engineering
%D 2014
%I Hindawi Publishing Corporation
%R 10.1155/2014/470201
%X The scope and responsibilities of control applications are increasing due to, for example, the emergence of industrial internet. To meet the challenge, model-driven development techniques have been in active research in the application domain. Simulations that have been traditionally used in the domain, however, have not yet been sufficiently integrated to model-driven control application development. In this paper, a model-driven development process that includes support for design-time simulations is complemented with support for simulating sequential control functions. The approach is implemented with open source tools and demonstrated by creating and simulating a control system model in closed-loop with a large and complex model of a paper industry process. 1. Introduction Model-driven development (MDD) is a system and software development methodology that emphasizes the use of models during the development work. In MDD, models conform to modeling languages that have formal metamodels, for example, unified modeling language (UML). In addition to manual development work, models can be processed with model transformations that revise existing and create new, refined models. The use of transformations may automate error-prone tasks such as importing information to models from preceding development phases and tools. Design models can be used for generating code or to analyze the developed systems. Automated model checks may reveal problems and inconsistencies in models and between phase products. The mentioned benefits of MDD are related to development tasks that are repetitive and simple enough to be treated with preprogrammed rules. However, MDD has not been able to, and probably cannot, automate all the complex tasks in system and software development. Demanding design decisions over alternative solutions to achieve (sometimes informal) objectives and product characteristics need to be made by professional developers. However, although genuine design decisions cannot be automated, developers do not always have to rely solely on their experience. For example, simulation is a technique that has been traditionally used in the domain within control algorithm development and control system testing. Automation and control system development is also an application domain in which the use of MDD techniques has been researched extensively during recent years. However, despite the research activities and the tradition of using simulations, ability to simulate early software design models has not yet been sufficiently addressed in the domain. In their previous
%U http://www.hindawi.com/journals/ase/2014/470201/