In recent years, wireless sensor and actor networks (WSANs) have become the subject of most studies about sensor networks because of their complexity and the vastness of the field. Keeping the connection between actors in WSANs is a matter of great importance. Failure of one or more actors can segregate the networks into several segments and consequently, stop the network operating properly. The aim of this paper is to propose a new method, named DGR, which utilizes the genetic algorithm to restore and keep the inter-actor connectivity of actors. After collecting the required data, some changes will be incorporated in the network architecture to reach an improved architecture for WSANs. DGR handles one possible case of multi-actor failure in two centralized and distributed stages. Using this proposed method and the effective localization of active nodes; more backup nodes will be achieved and restored with distributing processing efficiently. The obtained results from conducted simulations in the OMNet++ simulators demonstrate that DGR out performs Rim, DCR and DARA in the case of one failed node and RAM in the case of several failed nodes.