Handling System Complexity in Zonal E/E Architectures

• Published in: Transportation engineering (Oxford) 2023-09, Vol.13, p.100195, Article 100195
• Main Authors: Maier, Jonas, Reuss, Hans-Christian
• Format: Article
• Language: English
• Subjects: E/E architecture; k-means; System complexity; Vehicle powernet; Zonal architecture
• ISSN: 2666-691X; 2666-691X
• DOI: 10.1016/j.treng.2023.100195

•Zonal architectures further increase the complexity of automotive E/E architectures.•The UNICARagil platform is one of the first prototypes that has a zonal architecture•The presented method assists in the design process of zonal E/E architectures for vehicles and reduces their complexity•The method uses k-means clustering for the zone design, a genetic algorithm for the battery dimensioning and the dijksta algorithm for a holistic power supply design•A definition of complexity is given, which is based on components, interfaces and electric data As conventional electric/electronic (E/E) architectures reach their limits, new architecture concepts are being researched to make the complexity of the vehicle and its components manageable. Zonal architectures are a promising option, but there is a lack of methodology for developing them. This paper shows a method to design zonal E/E architectures, in particular their power supply. At first, the zonal vehicle power supply system (PSS) of the UNICARagil vehicle platform is presented. The UNICARagil project is a collaboration of 15 German research institutes and companies with the goal to develop disruptive modular architectures for agile automated vehicle concepts. To design a zonal, in regards of the vehicle functional architecture optimized PSS, it is necessary to cover the crucial aspects of both, the vehicle and the PSS, as well as their interfaces. A method is presented that optimizes the design of the PSS in zonal E/E architectures. The optimization is achieved by clustering the electric load positions to identify suitable positions for the zone control units (ZCU), wire harness routing by Dijkstra algorithm and integration of the entire PSS in vehicle packaging concepts. The aim is to reduce component cost, packaging restrictions and system complexity, which will result in a cut of development effort and a gain in user experience. The system complexity is therefor evaluated by metrics as well as in combination with other criteria by a simulation study.