The effective quantification and aggregation of the flexibility feasible domain for large-scale electric vehicles is a prerequisite for electric vehicle aggregators to participate in market bidding. However

the uncertainty of user travel information leads to a strong random characteristic in the adjustable capacity boundaries of aggregated electric vehicles

affecting the effectiveness of the aggregator's day-ahead bidding strategy. To address the randomness of user behavior and the uncertainty of frequency regulation prices

this paper proposes a method for reconstructing the feasible domain of electric vehicles and robust decision-making that accounts for the uncertainty of user travel information. First

we develop an inner approximation Minkowski reconstruction aggregation model that balances approximation accuracy and computational efficiency

incorporating uncertainties related to disconnection time and initial battery levels to characterize the robust boundary of the aggregation feasible domain. Second

we establish an ellipsoidal uncertainty set for frequency regulation prices and formulate a two-stage three-layer robust optimization model aimed at maximizing the aggregator's net revenue when participating in the joint market

and solve it using an improved column-and-constraint generation(C&CG) algorithm based on the extreme scenario method. Finally

through comparative analysis of numerical examples

we demonstrate the effectiveness and practicality of the proposed reconstruction aggregation method. The robust optimization model significantly mitigates the risks posed by real-world uncertainties

balancing revenue and risk

and guiding aggregators in formulating more reliable bidding strategies.