With the rapid growth in residential electricity demand

residential load flexibility is expected to become critical regulatory resources in power systems. Accurately modeling the bounded rationality of residential electricity consumption is essential for optimizing residential load-grid interaction decisions. This study categorizes electricity consumption into distinct mental accounts by analyzing the multi-dimensional needs of residential users and their correlations and substitutability. A multi-dimensional prospect function model of bounded rational electricity consumption behavior is constructed based on cumulative prospect theory and multi-attribute decision-making theory. Combined with mixed-Logit probabilistic modeling

a parameter identification method is proposed to address behavior uncertainty with heterogeneous data. On this basis

a decision-making method for residentialdemand response under bounded rationality is developed by integrating the mental account prospect functions

device power characteristics

and social requirements for grid interaction. Simulation results demonstrate that the proposed model can effectively characterize response preferences of residential load and confirm the effectiveness of the decision-making method in accounting for the bounded rational behaviors of residents.