Bayesian-DPOP for continuous distributed constraint optimization problems