One of the central approaches in the smart grid is to restore the smart distribution network (SDN) including different renewable sources and storage systems as microgrid formwork under fault conditions. Therefore, the automatic restoration processes need optimal power flow results in the SDN at fault times. Hence, this study proposed the optimal two‐stage coordinated energy management approach (CEMA) and self‐healing strategy (SHS) in the SDN in the presence of coupling neighbouring microgrids (CNMGs) according to autonomous decentralized restoration technique (ADRT). In the first stage, the SDN operation cost receiving from the upstream grid was minimized to obtain flexible and reliable network, followed by addressing network power flow equations, CNMG constraints including the renewable energy sources (RESs), energy storage systems (ESSs) and electric vehicles (EVs), and SDN operating limits. In the second stage, the optimal model of the ADRT formwork was defined based on the multi‐agent system (MAS). Also, the suitable restoration path was obtained with different zone agents distributed across the network CNMGs considering the minimization of the difference between the number of switching operation and priority loads restored as the normalized objective function. Furthermore, the scenario‐based stochastic programming was used to model the uncertainty of energy price, load, RES power, and EVs parameters. Finally, the proposed problem was solved by the Grey Wolf Optimization (GWO) algorithm and simulated on 33‐bus radial distribution network using MATLAB software. According to numerical results, the proposed strategy can offer the self‐healing and flexible‐reliable SDN in addition to the well‐suited overloading management for CNMGs.