Posicionamiento de vehículos inteligentes mediante la integración de GPS con sistemas inerciales (Análisis y diseño del sistema)

Nowadays, the GPS system is very common in all areas. It is also incorporated in some recent models of cars and smartphones too. The use of this device is growing rapidly because it has a great number of applications especially in navigation. The GPS technology is available worldwide 24 hours per day. It permits to locate the position of an automobile regardless of the time of use or weather conditions except if the GPS navigator suffers interferences which do not allow to receive the signal from satellites. As a consequence, this is a very accurate method to locate a car and, at the same time, it permits to improve the experience of driving. The space environment has between twenty-four and twenty-seven satellites which are located 20,000 kilometres away from Earth. A GPS navigator only needs the signal from four satellites to calculate its location with an error of less than twenty metres. It is possible to reduce this error to less than two centimetres if the signal is provided by five or more satellites. The main drawback is that GPS accuracy is not uniform depending on the place. There are factors which can reflect or block the signal, such as buildings, trees or tunnels. If this loss of signal stayed for a long period of time, navigation could turn out to be an impossible mission. An INS system can also locate the position of an automobile from a set of sensors which measure velocity and angular variations. This is an accurate system which is independent of external signals. However, the process depends on a correct initialization and the measurement corrupts with time. When they calculate positions independently, both systems have advantages and disadvantages, however it is possible to improve the resultant position when they are used together. In this thesis a data fusion method called Kalman Filter is the principal tool to improve GPS positioning. This method permits to incorporate the information received from both systems at the same time. The system itself could also decide the influence of each system to calculate the best result depending on the external situation. The results show that the Kalman Filter improves the navigation in good conditions, and it is also able to reduce bias when the GPS system is out of signal and permits to continue navigation with a smaller associated error