Demand response (DR) is considered to be an effective way to bring significant economic benefit to the commercial campus integrated energy system (CCIES) due to the large amount of flexible cooling and electric vehicle (EV) charging loads. To maximize DR's benefits

this paper proposes an integrated DR framework that includes direct load control for cooling loads and time-of-use for EV charging station load in the CCIES. Moreover

multiple uncertainties threaten the secure and economic operation of the CCIES. To deal with these challenges

this paper establishes an interval optimization (IO) based economic dispatch (ED) model

considering the uncertain parameters

including ambient temperature

DR parameters

pipeline parameters

and maximum available PV power output. To improve the solution efficiency

the nonlinear constraints are linearized by applying multi-layer perceptron and affine arithmetic. The order relation and the possibility degrees of intervals are used to transform the interval ED model into a bi-level optimization model. The extreme value theorem of linear interval functions is used to obtain the analytical expressions of the optimal solutions of inner-level models

and the ED model is finally transformed into a solvable mix-integer linear programming model. Test results from actual CCIES demonstrate that the DR can improve the economy and reduce the uncertain fluctuation range of both the objective function and state variables. The ED result can maintain an economical and secure operation under multiple uncertain fluctuations.