Abstract: To address the pronounced problems of long regulation time lag and severe indoor temperature unevenness in central heating, the central-heating peak-shaving systems achieves regulation optimization through a combined heating mode of basic central-heating guarantee and fluctuating supplement by the peak shaving system, and has become one of the prevailing heating configurations in northern China. Meanwhile, the central-heating peak-shaving system coupled with distributed green power can facilitate building–grid interaction, enhance renewable power accommodation, and satisfy residents' personalized thermal comfort demands. Therefore, this paper proposes a central-heating peak-shaving system coupled with distributed green power and oriented toward personalized heat demands, and incorporates the bounded rationality of residents participating in central-heating demand response into the modeling. A cost–comfort model is employed to characterize the demand-response behaviors of heterogeneous users, and a bi-level pricing and scheduling optimization framework for heating enterprises and users is established to achieve coordinated optimization of green power accommodation and heating benefits. Taking a residential community in Xi'an for case study, the results show that the daily average accommodated green power reaches 10~18 MW, the operating cost of the heating enterprises is reduced by 36.5%, and the user temperature compliance rate is increased from 2.9% to 90.3%. The proposed approach can alleviate the scheduling deviation caused by the idealized assumption of residents' demand response, improve system economy and terminal comfort guarantee level, and provide a reference for the participation of green power in heating peak shaving and the design of differentiated pricing mechanisms.