Following a massive outage

it is essential for smooth network reconstruction to identify a safe and effective units’ restoration sequence. Based on the definition of unit’s restoring risk and taking the duration into consideration

the concept of risk of electrical energy loss during the network reconfiguration is further proposed. Through integration of crossover particle swarm optimization and Floyd algorithm

the risk of electrical energy loss can be minimized by means of detailed simulation of network reconfiguration. As a result

the optimal units’ restoration sequence with higher operational reliability as well as the best technical performance is resolved. In particular

investigation on the effect of the unit’s off-line duration on unit’s starting and reconfiguration progress is made closer to field situation by calling the embedded simulation program of thermal unit’s starting and load picking-up. The outcome of its application on the New England 10-unit 39-bus power system indicates the validity of the method proposed. This work is supported by National Natural Science Foundation of China (No. 51077052).