FPGA implementation of neuro-fuzzy system with improved PSO learning. (July 2016)

• Record Type: Journal Article
• Authors: Karakuzu, Cihan; Karakaya, Fuat; Çavuşlu, Mehmet Ali
• Is Part Of: Neural networks. Volume 79(2016)
• Page Start: 128
• Page End: 140
• Publication Date: 2016-07
• Subjects: Neuro-fuzzy network -- FPGA implementation -- VHDL -- System identification -- Metaheuristic learning; Neural computers -- Periodicals; Neural networks (Computer science) -- Periodicals; Neural networks (Neurobiology) -- Periodicals; Nervous System -- Periodicals; Ordinateurs neuronaux -- Périodiques; Réseaux neuronaux (Informatique) -- Périodiques; Réseaux neuronaux (Neurobiologie) -- Périodiques; Neural computers; Neural networks (Computer science); Neural networks (Neurobiology); Periodicals; 006.32
• DOI: 10.1016/j.neunet.2016.02.004
• Languages: English
• ISSNs: 0893-6080

Abstract

Abstract: This paper presents the first hardware implementation of neuro-fuzzy system (NFS) with its metaheuristic learning ability on field programmable gate array (FPGA). Metaheuristic learning of NFS for all of its parameters is accomplished by using the improved particle swarm optimization (iPSO). As a second novelty, a new functional approach, which does not require any memory and multiplier usage, is proposed for the Gaussian membership functions of NFS. NFS and its learning using iPSO are implemented on Xilinx Virtex5 xc5vlx110-3ff1153 and efficiency of the proposed implementation tested on two dynamic system identification problems and licence plate detection problem as a practical application. Results indicate that proposed NFS implementation and membership function approximation is as effective as the other approaches available in the literature but requires less hardware resources. Highlights: A novel approach for implementation of neuro-fuzzy system on FPGA is proposed. Hardware implementation includes meta-heuristic learning of the system. A novel membership function implementation approach is also presented. The proposed approaches have been experimentally tested for both benchmark and practical problems. The obtained results show that the proposed method is effective and acceptable in handling various types of real life problems.