Driven by modern advanced informationand communication technologies, distributed energy resources have greatpotential for energy supply within the framework of the virtual power plant(VPP). Meanwhile, demand response (DR) is becoming increasingly importantfor enhancing the VPP operation and mitigating the risks associated with thefluctuation of renewable energy resources (RESs). In this paper, we proposean incentive-based DR program for the VPP to minimize the deviation penaltyfrom participating in the power market. The Markov decision process (MDP)with unknown transition probability is constructed from the VPP'sprospective to formulate an incentive-based DR program, in which therandomness of consumer behavior and RES generation are taken intoconsideration. Furthermore, a value function of prospect theory (PT) isdeveloped to characterize consumer's risk attitude and describe thepsychological factors. A model-free deep reinforcement learning (DRL)-basedapproach is proposed to deal with the randomness existing in the model andadaptively determine the optimal DR pricing strategy for the VPP, withoutrequiring any system model information. Finally, the results of cases testeddemonstrate the effectiveness of the proposed approach.