Design and development of a test instrument for steering wheel characterization in automotive safety tests. Test and analysis data for different algorithms validation

This thesis activity is aimed at designing and building a prototype of an instrument for measuring the steering wheel angle and the steering wheel torque, with the vehicle stationary and in motion. The project has been carried out at the Turin-based company AMET Srl. The instrument prototype, called Electronic Level, must have an appropriate mechanical fixing system able to adapt to different types of steering wheels, it has intelligence onboard programmed for carrying out tests, storing data, processing CAN messages, and providing a PC interface for offline data processing. The proposed system consists of three components: the first one is the Electronic Level fixed to the steering wheel in three or four points to guarantee a secure and deterministic fixing, avoiding damages to the steering wheel and avoiding impediments regarding safety systems of the vehicle as airbags. The second component is an OBD (On-board Diagnostic) interface for the acquisition of steering angle, steering torque, when available, vehicle speed, and VIN (Vehicle Identification Number) directly from the CAN bus of the vehicle under testing and able to send all these information to the Electronic Level. These two components communicate to each other through Wi-Fi communication: a custom protocol has been implemented for this purpose. The last system component is the software to be loaded on the ground station (i.e. the PC), equipped with a USB port to interfacing with Electronic Level. During the preliminary analysis phase, it has been decided to use a 9-axis IMU system, which includes accelerometer, gyroscope, and magnetometer sensors, that allow measuring the steering wheel angle, employing a specific algorithm. The idea is to filter the values obtained from these sensors using adaptive filtering techniques, such as Complementary Filter or Extended Kalman Filter, to compensate measure errors. Due to the lack of time, it has not been possible to realize an effective solution for the torque measurement, although a preliminary feasibility study is present, so it can be used in future developments. The obtained results have been examined for analyzing the accuracy of the steering wheel angle measurements, in different tests condition. For future developments, in addition to the implementation of the torque estimation, an effective anchoring system and a better data filtering method based on the Extended Kalman Filter should be evaluated.