It is acknowledged that the affective cognition of decision-makers plays an important role during the course of multi-objective decision-making. Aiming to enhance the decision-making ability of computers by enabling them with human’s affective cognition, layered affective model is proposed, which merges personality, mood and emotional states and is able to describe the quantified affective cognition of decision-makers in solving complex decision-making problems. This layered affective cognitive model can be applied to the creation of virtual human decision-makers incorporated with human’s perceptual and rational characteristics. Instead of human decision-makers, the virtual ones participate in solving multi-objective optimization problems by means of interactive decision-making approaches. Numerical examples are employed to demonstrate the effectiveness of the proposed model.