Constraint programming is a successful technology for solving combinatorial problems modeled as constraint satisfaction problems (CSPs). An important extension of constraint technology involves problems that undergo changes that may invalidate the current solution. These problems are called Dynamic Constraint Satisfaction Problems (DynCSP). Many works on dynamic problems sought methods for finding new solutions. In this paper, we focus our attention on studying the robustness of solutions in DynCSPs. Thus, most robust solutions will be able to absorb changes in the constraints of the problem. To this end, we label each constraint with two parameters that measure the degree of dynamism and the quantity of change. Thus, we randomly generate a set of more restricted CSPs by using these labels. The solutions that satisfy more random CSPs will have a higher probability of remain valid under constraints changes of the original CSP.
展开▼
