Hardware-in-the-Loop test bench and Battery Model design for a Hybrid Electric Lamborghini prototype

Hardware-in-the-Loop (HIL) simulation is a real-time simulation that allows for testing of embedded codes without resorting to a real hardware system but only a hardware under test, also called Device Under Test (DUT). This provides the ability to test for abnormal conditions and errors that could damage the hardware if the code performs differently than expected. Control systems are fundamental in different fields, and verifying an embedded code is complex due to the risk of hardware damage at an advanced stage compared to that of prototyping when making changes becomes complex. HIL simulation offers an ideal solution for testing embedded codes before they are implemented on real hardware systems, as it allows for testing of extreme scenarios that could damage or even destroy the hardware under test. The performance of HIL simulation depends on both the complexity system’s model and the real-time machine adopted. For example, the dynamics of a motor or an inverter can be changed by varying how the power electronics are modeled and the impact this will have on the computer processor used for HIL must be considered. The performance also influences the real-time behavior capability of the system. For simulations at 1 or more MHz, the real-time system must have low I/O latency and a processor capable of reaching the operational frequency. The choice between CPU and FPGA for simulating systems also depend on the complexity of the model, the mathematical details of the system, and the number and types of I/O used during the testing phase. It becomes clear that the design of a HIL bench is a compromise between the accuracy of the simulation, the ability to perform simulations in real time and the characteristics of the components and processors used for the bench witch consequently increase the related costs. A HIL test begins with a MATLAB and Simulink simulation, which allows for quick and accurate full simulation in real-time. Complex system models can then be created, and a variety of scenarios tested, including control system, battery pack and load models. MATLAB and Simulink simulation can help to identify any anomalies present in the embedded code before performing a real simulation. Additionally, a HIL simulation can be run multiple times to ensure that the embedded code is functioning properly and can be used to test the scalability of the embedded code across larger and more complex systems. The primary goal of this project is to design a HIL bench capable of simulate a complete vehicle, not for testing or validation, but rather to analyze components and control units to investigate potential problems that may arise in production vehicles. By doing so, it will be possible to easily replicate the scenario that caused the error and test the same component multiple times in the same environment, allowing for comparison with a different specimen. Taking into consideration the complexity of the project, it was decided to split it into three distinct parts in order to ensure the best possible outcome. The details and reasoning behind these choices will be discussed in depth, with the aim of understanding the motivations and benefits of this approach. Particular focus will be given to the initial phase, where all elements will be thoroughly examined, with a special emphasis on the design of the battery model required for the bench's operation. This is a critical aspect, as the model must be able to provide the necessary trade-off characteristic for the bench to operate efficiently.