Knowledge Based Engineering to support electric and electronic system design and automatic control software development

The level of software integration in vehicles, such as electric cars and aircraft, is rapidly increasing. Due to the increasing complexity of the embedded control software, significant delays can occur in development programs or even errors can be present in the control software of the final product. In the development of the electric and electronic system (E/E system), the analysis and specification for the architecture of the logical system, technical system and software itself includes many repetitive (manual) processes. Those repetitive processes are time consuming and are prone to errors. This research proposes new methods and tools that allow the designer to take electronic components, including control software, into account already in the conceptual design stage of complex systems. These new methods are based upon the principles of Knowledge Based Engineering (KBE), which is essentially a combination of computer aided design (CAD) and artificial intelligence (AI). The proposed methods can establish the relationship from the logical system architecture to the technical system architecture and finally the software components. Moreover, the proposed tools can model the logical, technical architecture and automatically generate the software components. The software language GDL, which is particularly suitable for the representation of complex systems and the development of KBE applications, has been used to develop the tools. The development of an Anti-lock Braking System (ABS) for a novel electric vehicle configuration has been chosen as test case. Based on a single intelligent product model, that contains the main design parameters of the vehicle specified by the designer, two models are generated automatically; (1) the simulation model of the physical plant and the associated control system, and (2) the control software. For a specific vehicle configuration, the simulation model can be used to test the control system and to optimize the parameters of the control system. In the case of an ABS, a braking maneuver is simulated. Next, the software components are generated automatically. The simulation model is used to test the software components for a range of conditions. The results show that the of the software components are automatically updated when the physical plant of the E/E systems or top level overall design changes. The final source code is well-structured and easy to understand due to the fact that there is a direct relation between the vehicle design parameters specified in the original product model and the variables and their values in the data model of the software components. The proposed design methods and tools can in principle be applied to any dynamic system with a high level of software integration, such as e.g. unmanned aerial vehicles.