Real-time fuzzy processor on a DSP

Fuzzy logic has been successfully introduced in control applications, but usually has the problem of low speed. Fuzzy logic applications can be seen as any other signal processing application, which makes a DSP a good candidate for high speed fuzzy processing, taking also into account that DSPs are classical platforms in industrial environments. This paper presents a real-time full-programmable fuzzy processor using piecewise-linear interpolation techniques and implements it using a DSP. A full-programmable fuzzy processor is defined as a system where the t-norm, t-conorm aggregation operator, propagation operator, rules, membership functions and defuzzification algorithm can be defined by any valid algorithm. Real-Time fuzzy processing is defined as a way of processing the knowledge base in a constant time and with a minimum speed of 1 MFLIPS.

[1]  Masaki Togai,et al.  Expert System on a Chip: An Engine for Real-Time Approximate Reasoning , 1986, IEEE Expert.

[2]  Yan Shi,et al.  Fuzzy inference chip FZP-0401A based on interpolation algorithm , 1998, Fuzzy Sets Syst..

[3]  Herbert Eichfeld,et al.  A 12b general-purpose fuzzy logic controller chip , 1996, IEEE Trans. Fuzzy Syst..

[4]  C. von Altrock,et al.  Recent successful fuzzy logic applications in industrial automation , 1996 .

[5]  Serge Boverie,et al.  Applications of fuzzy logic: towards high machine intelligence quotient systems , 1997 .

[6]  R. Rovatti,et al.  Linear and fuzzy piecewise-linear signal processing with an extended DSP architecture , 1998, 1998 IEEE International Conference on Fuzzy Systems Proceedings. IEEE World Congress on Computational Intelligence (Cat. No.98CH36228).