Problème de satisfiabilité booléenne (SAT)
The Boolean satisfiability problem (often abbreviated as SAT) is a fundamental problem in l'informatique and logique mathématique. It involves determining whether a given Boolean formula can be satisfied by some assignment of truth values (true or false) to its variables. In simpler terms, SAT asks the question: can we make the entire formula true by choosing appropriate values for its variables?
Une formule booléenne typique est exprimée en forme conjonctive normale (CNF), which is a conjunction (AND) of clauses, where each clause is a disjunction (OR) of literals (variables or their negations). For example, the formula (A OR NOT B) AND (B OR C) is in CNF.
Finding a satisfying assignment is crucial because many problems in computer science can be framed as SAT problems, especially in fields like intelligence artificielle, verification, and optimization. The significance of SAT lies not only in its theoretical importance but also in its practical applications, such as in circuit design, software testing, and automated reasoning.
SAT is classified as NP-complete, meaning that while it is easy to verify a solution (i.e., checking if a particular assignment satisfies the formula), finding a solution can be computationally challenging. The development of efficient algorithms and heuristics, such as the DPLL algorithm and modern SAT solvers, has made it possible to tackle large and complex SAT instances effectively.
En résumé, le problème de satisfiabilité booléenne consiste à déterminer l'existence d'assignations de vérité qui peuvent rendre une formule logique vraie, servant de pierre angulaire dans diverses applications du calcul et de la logique.