With the wide spread of smart meter and smart apparatus， load control has become an important tool for system operation and control. As for the service restoration problem， the controllable load can be a new consideration in achieving the optimization objectives. A new optimization model with the interruptible load taken into account is developed for the service restoration problem. The incorporation of the interruptible load is helpful to not only reducing the amount of curtailment of priority load due to inadequate capacity to restore the whole out-of-service area， but also improving the reliability of supply by decreasing the switching operation time given the availability of the numerous manual switches in the distribution system. The service restoration problem is a constrained multi-objective optimization problem that uses many algorithms to solve this so-called NP-complete problem. This paper employs the NSGAA-Ⅱ algorithm to solve the model developed. With the NSGA-Ⅱ algorithm， it is possible to obtain the optimum solution set of each objective while helping the operator to choose the most appropriate plan for service restoration. Application of the model proposed and the NSGAA-Ⅱ algorithm to the IEEE 69 distribution system has verified their feasibility and correctness.