Home   |   Members only   |   Contact   

Site menu:

Location: mbse4mechatronics.org / About 

About

MBSE4Mechatronics:Background - Objectives


Background

As a result of the increasing complexity of modern mechatronic systems, designers are no longer able to comprehend all design aspects during the design process without computational support. Therefore, this project aims to develop a model based systems engineering (MBSE) methodology for mechatronic systems, which will guide the designer through the design. By integrating information from various disciplines during all stages of the design, optimised products will be realised in a fast and efficient way. Although a model based design approach is now a well accepted practice in individual engineering disciplines, a plethora of new challenges arise when such an approach is applied to multi-disciplinary mechatronic design. Making optimal cross-disciplinary design decisions and guaranteeing consistency between discipline-specific developments are two main challenges. To tackle these challenges, the MBSE4Mechatronics project aims to develop MBSE techniques to:

  1. Support the decision making process of system architect(s), who integrates discipline-specific developments, during the early stages of the design when the system architecture is fixed;
  2. Support automatic synchronization of discipline-specific developments, especially in the later design stages.

The core idea of the MBSE4Mechatronics approach to realize these objectives, is the use of a central model of the system architecture throughout the complete design process. This model, that will serve as a pivot model to coordinate the discipline-specific developments, will be gradually refined during the design process. To better support the decision making process of the system architect (objective 1), the MBSE4Mechatronics project will derive a central multi-disciplinary system architectural model as well as multiple discipline-specific analysis models, which can be used to evaluate the feasibility and optimality of a system architecture.

  1. The system architectural model will incorporate a formal representation of the feasible design space, which can be used to check whether design constraints are fulfilled. To this end, methods will be developed to formally represent the design space of feasible architectures in terms of the design parameters.
  2. The multiple discipline-specific analysis models will be integrated, which will allow assessing the overall system behavior and performance. In order to keep the level of complexity of the analysis models reasonably low, the detailed models which are used in the individual engineering disciplines, must be simplified to a higher abstraction level. Also the advanced design approaches, which are followed in individual engineering disciplines, must be simplified to design rules, which can be understood and applied by a system architect. To this end, methods will be developed to
    • derive multi-domain analysis models at a high level of abstraction from existing detailed analysis models
    • create simplified design rules that can be used by the system architect when updating the system architecture.

Taking this one step further, (semi-)automatic exploration algorithms, which generate new system architectures in an intelligent way, will be developed to guarantee that the optimal system architecture is identified. Therefore:

  1. a formal description which allows to represent the requirements and objectives in a mechatronic design process will be realized;
  2. intelligent design space exploration techniques will be extended to make them applicable for multi- disciplinary mechatronic problems.

To ensure automatic synchronization between discipline-specific developments (objective 2), a model based infrastructure will be created that allows to define declarative relationships between semantically related objects in different models. A key element in this infrastructure will be a central pivot model repository. This repository should consist of discipline-specific meta-model information as well as structural and parametric links between the elements in these meta-models. The developed infrastructure will be applied to evaluate and enforce consistency between the different discipline-specific models. Finally, to ensure the industrial relevance of the project, all the work packages will demonstrate their results on industrial cases.

Top