In applications where low cost is the primary concern, analog VLSI controllers are an option. In this paper a hardware and software integrated toolkit for development of analog controllers based on fuzzy logic is presented. The system is composed of a CMOS analog circuit whose input/output relationship can be programmed to synthesize various control functions. The value of the programmable parameters can be automatically computed by the software tool, assuming that the user specifies the controller expected behavior by means of linguistic and/or numeric information. A complete example of how fuzzy linguistic and numerical methodologies can be combined with more conventional techniques to synthesize a non-trivial controller is detailed down to measured data. A final discussion clarifies the main advantages of the fuzzy analog computation when it is supported with an appropriate software layer: non-trivial solutions to problems in which accuracy is not the primary concern can be easily synthesized and integrated into very small silicon areas with small power requirements.
|Microelectronics Group Home Page||Staff|
|Research Activities||Teaching Activity|
|DEIS Home Page||University of Bologna Home Page|