Device positioning algorithm for an automotive application

This thesis presents an algorithm that calculates the mounting angles for a black box device installed on a car. When mounting the device on a car, technicians could use specific machines to install it. But this process introduces errors and the devices would not be installed the same way on different cars. This results in incorrect information coming from the car. Moreover, the installation process requires more attention and so it results more difficult and expensive. To solve this problem and obtain correct information from the car, the axes of the sensor inside the device have to be aligned with the axes of the car. The device contains a three-axis accelerometer and a GPS module. The roll, pitch and yaw angles are estimated by using a data fusion algorithm. Several algorithms have been developed to find the most accurate results for the mounting angles. These algorithms were tested with data collected on paths with different road characteris- tics. In this thesis it is proposed a solution that gives the most accurate results, with some observations. The solution has been developed by using the model-based software design approach. Model-in-the-loop testing has been performed to refine the implemented model and to improve it by an offline elaboration of the actual data produced by the device modules. The model has been translated in C code, which was tested by performing Software- in-the-loop testing. Finally, the code has been uploaded on the device and properly tested by mounting the upgraded device, which was positioned in several ways, on the car.