Estimation and Sensor Fusion techniques on an autonomous vehicle

A model is a simplified or partial representation of reality, defined in order to accomplish a task or to reach an agreement. It is based on an original system reflecting the relevant selection of the original system properties. The model is the central artifact of software development, it includes the documentation, static analysis, rapid prototyping, automated testing, refactoring, transformation and code generation. The increasing complexity of software in engineering applications leads to problems in the software development like production delay, wrong functionalities, software poorly documented or commented. Modelling allows to increase productivity, efficiency, reusability, improve portability of the software, easy to test. This are the main reason why nowadays the model-based method is being used in the complex engineering projects. In this thesis it will be discussed specifically the model of a skid steering mobile robot, courtesy of Brain Technologies, with ultrasonic sensors in order to define its position with respect another object and follow the tracked object in an S shape path by means of Matlab and Simulink. In the end, after testing the single components and the whole model of the robot, the model will be code generated via Embedded coder and deployed in the Arduino board. The code generated by Simulink used a huge amount of memory and it was needed to reduce and remove some redundant code generated.