Hardware oriented fuzzy neural network
Boubekeur Mendil, K. Benmahammed · 2002
In this paper, a network of fuzzy neurons is proposed. Fuzzy neurons are obtained by introducing a slight modification on the formal model of the traditional neurons. The major advantages of this architecture is that it can use both numerical and linguistic knowledge. Furthermore, the network consists of identical nodes. This greatly simplifies the hardware implementation. In order to demonstrate the effectiveness of the proposed architecture, we have compared its performance with that of fuzzy logic using the truck-backing up example.