Neuro-fuzzy based autonomous vehicles

One way of dealing with the real world phenomena is qualitative and non-numerical in nature. In decision making processes as in the guidance of vehicles, masses of numerical data are converted into some qualitative form and thus are dealt with only in aggregation, e.g. visual perception-steering activity. This form of aggregation gives rise to a set of linguistic labels and is sometimes referred to as information granules. This aggregation of information makes the partition of space more manageable for further processing. All cognition and inference processing are then carried out at the level of the granules. This process of aggregation or granulation implies that we deal with the relationships of functions between linguistic label rather than with numerical quantities. To cope with this style of cognition, a suitable modelling technique is developed using the theory of fuzzy sets since this theory deals with such granularity of our perception.<
>