Software in the Loop: integration between vECU and plant model

This thesis is related on a collaborative project between Bosch (RB) and Iveco, aimed at creating the Software in the Loop (SiL) environment. The new simulation environment will offer increased flexibility and speed in testing and simulations compared to the current working model. RB's main tasks involve the development of hardware (HW), software (SW), and datasets for the engine control unit (ECU), which are rigorously tested before delivery. The objective of the SiL project is to create a flexible simulation environment that can accurately replicate HiL test results. RB's responsibilities include virtualizing the ECU and creating an interface between it and the plant model, which is provided by the customer. The development of this new simulation environment commenced in September, with an estimated completion by mid-2025, and the first release planned for March. The thesis focuses on the first release, which is a simulation environment developed in open loop, with a standalone virtual engine control unit (vECU). Creating the interface between these two realms proved challenging due to the complex customer-modeled system, which simulates the plant model, and due to the intricacy of the SW structure of the engine ECU. The thesis explains the intricate task of associating input/output (IO) signals of the plant model and vECU. This is accomplished through a virtual Electric Layer (vEL), which also simulates the behavior of devices connected to the engine control unit, such as sensors and actuators. IO calibration is performed using cdfx files generated by the INCA tool of ETAS. RB also supported the customer by creating models for certain H-Bridge actuators (e.g., EGR Valve) using a Neural network approach. The first chapter provides an overview of Bosch's work on engine ECUs for customers, followed by an explanation of the SiL project's releases. The second chapter discusses the software structure of the physical engine ECU developed by Bosch, with a focus on the device library, and signal sensing and conditioning functionalities. The third chapter is related to the artifacts of the SIL environment, explaining the tasks and the way of communication. In this chapter, the different development choices are explained in particular on the different types of vECU. The fourth chapter is related to my work, starting from the study of the SW documentation, passing through the association of signals and calibration of the devices connected to engine control unit. At the end of the chapter it is explained the creation of the EGR valve model using the neural network approach. The last chapter is on the results of the first release of the vECU, trying to create a discussion on the enhancement that can be done in order to replicate well the tests on the Hardware in the Loop.