Aiming at the economic stability of substation integrated energy system, a multi-time scale optimization scheduling strategy based on model predictive control（MPC）is proposed, which combines ground source heat pump and demand response. Firstly, the system architecture is constructed and the MPC algorithm flow is described. In day-ahead scheduling, by considering the randomness of TOU and new energy, an optimization model aiming at minimizing operation cost is established, and the optimal output and energy storage plan of the unit are determined. Intra-day scheduling uses MPC algorithm to correct the day-ahead schedule in real time to reduce the impact of uncertainty on the power grid and improve the economy. The demand response characteristics and energy efficiency level of ground source heat pump are discussed by comparing different schemes through example analysis. The results show that the proposed model can optimize the system operation under the demand response environment, improve the equipment utilization rate, reduce the operating cost, and promote the consumption of renewable energy.