Solving Dichotomy-based Constrained Encoding With Twin Graph Coloring

Constrained encoding consists of finding a minimum length state encoding that meets some requirements. Dichotomy-based constrained encoding is more general than other constrained encoding frameworks, but it is also more difficult to solve in practice. This paper introduces a new formalization of this problem, which reduces it to what we call a twin graph coloring. This new problem leads to original exact and heuristic algorithms. Experimental results show that the exact solver is more efficient than the best known exact solvers, and that the heuristic version is competitive.