Method and architecture for fuzzy-logic using unary processing
OwnerMohammadhassan Najafi, Amir Hossein Jalilvand +1inventors
InventorsMohammadhassan Najafi, Amir Hossein Jalilvand, Mahdi Fazeli
Life of this patent
Estimating the term…
Abstract
Efficient hardware design of the fuzzy-inference engine has become necessary for high-performance applications. The disclosed technology applies unary processing to the platform of fuzzy-logic. To mitigate the latency, the proposed design processes right-aligned bit-streams. A one-hot decoder is used for fast detection of the bit-stream with maximum value. Implementing a fuzzy-inference engine with 81 fuzzy-inference rules, the disclosed architecture provides 82%, 46%, and 67% saving in the hardware area, power and energy consumption, respectively, and 94% reduction in the number of used LUTs compared to conventional binary implementation.
Claims10
An architecture for performing fuzzy interference in unary computing systems, comprising:
- a fuzzification module;
- a unary number generator;
- a rule-apply module;
- an aggregation module;
- and
- a defuzzification module, comprising:
- one or more inputs comprising aggregated values;
- a database comprising at least one prestored value;
- a one-hot decoder;
- and
- a MUX unit;
- wherein the aggregated values comprise one or more outputs from the aggregation module;
- wherein the aggregated values further comprise corresponding unary bit-streams;
- and
- wherein the one-hot decoder comprises functionality to detect which unary bit-stream has the maximum value.
A method for performing fuzzy interference in unary computing systems, comprising:
- a. providing an architecture, comprising:
- a fuzzification module;
- a unary number generator
- a rule-apply module;
- an aggregation module;
- and
- a defuzzification module;
- b. performing a fuzzification process by the fuzzification module, comprising converting one or more crisp inputs to one or more fuzzy linguistic values;
- c. converting numbers in a range of [0,1] of the fuzzy linguistic values to unary bit-streams by the unary number generator;
- d. applying one or more inference rules by the rule-apply module;
- e. aggregating one or more outputs of the rule-apply module;
- and
- f. performing a defuzzification process by the defuzzification module, wherein the fuzzy linguistic values are converted to one or more crisp value outputs.
Dependencies are inferred from the claim text.