Abstract: After an outage in extreme hot weather, the flexibility of air conditioners is used to formulate a load restoration scheme to fit the change of load amount, available power and outdoor temperature, which can restore load safely and reliably and effectively enhance power system resilience. In this regard, considering the active control of air conditioners, a multi-time scale substation load restoration decision-making method is proposed. First, a framework of multi-time scale load restoration is formulated to coordinate the long-time scale feeder switch scheme making and short-time scale active air conditioners control modulation. Second, a long-time scale feeder switch scheme optimization model is established to make feeder switch schemes to improve load restoration efficiency from the perspective of whole process of substation load restoration；Model predictive control is adopted to develop a short-time scale air conditioners control modulation to fit the continuously changed load amount, available power and outdoor environment during restoration, taking the minimum impact of customer comfort. The simulation results demonstrate that the proposed substation load restoration decision-making method can make schemes adaptively with better restoration benefit under extreme hot weather.