Smart home, with the Internet of Things as the core, will substantially change the time scale of residents’ electricity consumption, increase the flexibility of power load, and provide more opportunity for demand response and related services, but also bring huge uncertainties. Focus on the transferable function of smart home working time, we are committed to exploring the possible changes of smart home participating in demand response. Here, we establish the smart-home integrated management model with goals of minimizing the electricity cost and peak-valley difference, and provide an optimization scheme that integrates smart home into demand response. Furthermore, the peak shaving potential of smart home participating in demand response and its impacts on power supply and grid investment are evaluated. Our results show that the time-of-use policy can reduce the peak-valley difference between -29.2% and -23.9%, and cut electricity cost by up to 9.5%, which is conducive to encouraging smart home to participate in demand response. The shift of working time in smart home may increase power consumption and increase residential electricity costs by 4.8%-11.5%. In the unconstrained dispatch mode, driven by smart home participation in demand response, it is expected to reduce the peak load of the power grid by 141 to 149 million kilowatts, and reduce power supply and power grid investment by 1.13-1.19 trillion yuan.