Autonomous-Centered Problem Allocation Oriented to Cooperation

Abstract

By reasonably allocating a cooperative problem which need multiple solvers cope with together, the problem could be performed more effectively and efficiently. A problem could be divided into multiple sub-problems; each has certain ability requirement which is the hinge to relate problem and solver. According to ability requirement, the solver candidate set for each sub-problem could be established. To select suitable solver from candidate set so as to solve a cooperative problem in more autonomous and consistent way, a mathematical allocation model taking the minimization of interaction number as objective function is established. The model solving process is deployed by decreasing two kinds of interactions, i.e. intra-interaction and extra-interaction. Experiment shows this method obtains better performance than general allocation.