Differential Evolution Based on Estimation of Distribution for C-DCOP

Aiming at the shortcomings of the existing Continuous Distributed Constraint Optimization Problems (C-DCOP) solving algorithms, such as the weak search ability, the inability to guarantee convergence, and the solution quality to be improved, Differential Evolution based on Estimation of Distribution for C-DCOP (DEED) is proposed, and the elite strategy and the reservation strategy guarantee the the convergence of DEED, the solution quality is effectively improved by introducing the estimation of distribution, in addition, the adaptive crossover probability enhances the search ability. Finally, the experimental results on four types of benchmark problems show that compared with the state-of-the-art C-DCOP solving algorithms, DEED exhibits excellent performance in terms of solution quality.