A high performance fuzzy logic architecture for UAV decision making

The majority of Unmanned Aerial Vehicles (UAVs) in operation today are not truly autonomous, but are instead reliant on a remote human pilot. A high degree of autonomy can provide many advantages in terms of cost, operational resources and safety. However, one of the challenges involved in achieving autonomy is that of replicating the reasoning and decision making capabilities of a human pilot. One candidate method for providing this decision making capability is fuzzy logic. In this role, the fuzzy system must satisfy real-time constraints, process large quantities of data and relate to large knowledge bases. Consequently, there is a need for a generic, high performance fuzzy computation platform for UAV applications. Based on Lees’ [1] original work, a high performance fuzzy processing architecture, implemented in Field Programmable Gate Arrays (FPGAs), has been developed and is shown to outclass the performance of existing fuzzy processors.