When solving the service restoring（SR） problem using the intelligent optimization algorithm， considerable time has to be consumed and the optimizing performance of the algorithm is reduced as it is necessary to not only repeatedly do network topology analysis during changes of the network structure but also very often to do additional checking on radiation during optimization to ensure that the constraints on network radiation operation are not violated. To avoid these disadvantages， the node-depth encoding（NDE） technique is introduced into the non-dominating sorting genetic algorithm-Ⅱ（NSGA-Ⅱ） to solve the SR problem to replace traditional crossover and mutation operations with preserve ancestor operator（PAO） and change ancestor operator（CAO） operations in NDE technique. The topological structure of the system can be quickly obtained with the NDE technique. Simulation results show that the algorithm proposed has a better performance in convergence and diversity than NSGA-Ⅱ.