Power supply service represents the final stage in delivering electricity to end-users, and enhancing its capability contributes to improved power stability for consumers. Fault repair is the most common power supply service task, with equipment failure being the primary focus during service operations. This paper proposes a two-stage emergency repair optimization scheduling approach that integrates repair resources and task optimization. Addressing issues such as uneven allocation of emergency resources, the model considers the assignment of emergency power generation vehicles to fault points, resource types, resource quantity variability, and practical emergency repair scenarios under large-scale power failures caused by emergency incidents. Innovatively, a multi-objective distribution network repair model is developed, incorporating repair modes and resource constraints, with the primary objective of minimizing repair time and the secondary objective of reducing economic losses. Furthermore, a multi-mode model for distribution network repair task scheduling is established, along with a two-stage scheduling method. The first stage optimizes the allocation of power generation vehicle resources based on fault locations, while the second stage optimizes repair task scheduling by considering both repair time and socio-economic losses. A serial scheduling generation mechanism is employed to generate repair task scheduling schemes, and an improved moth-flame optimization algorithm is applied to obtain the optimal solution for the repair task scheduling model.