// Rules: IF temperature IS warm THEN fan_speed IS medium engine.addRule("temperature", "warm", "fan_speed", "medium"); engine.addRule("temperature", "hot", "fan_speed", "high");
// Output variable "fan_speed" engine.addVariable("fan_speed"); engine.addFuzzySet("fan_speed", "low", afl::Triangular(0, 0, 50)); engine.addFuzzySet("fan_speed", "medium", afl::Triangular(30, 50, 80)); engine.addFuzzySet("fan_speed", "high", afl::Triangular(70, 100, 100)); afl library
Its associative design offers unmatched flexibility, but that same design limits performance and type safety. It is not suitable for large-scale, high-performance, or safety-critical applications. // Rules: IF temperature IS warm THEN fan_speed
| Class | Description | |-------|-------------| | FuzzySet | Represents a membership function (triangular, trapezoidal, Gaussian). Contains parameters (a,b,c,d) and methods to compute membership degree. | | LinguisticVariable | Holds multiple FuzzySets (e.g., "Temperature" with sets: Cold, Warm, Hot). Uses std::map<string, FuzzySet> . | | FuzzyRule | IF-THEN rule with antecedent and consequent (both are std::map<string, double> or strings). | | FuzzyEngine | Main inference engine. Stores linguistic variables and rules. Performs fuzzification, inference (Mamdani), and defuzzification (Centroid, Bisector, etc.). | | | FuzzyRule | IF-THEN rule with antecedent
Usually MIT or BSD – permissive. 2. Architecture & Core Components The library is structured around four main classes (all in namespace afl ):