A switching control strategy is proposed for improving the economic performance of the economics-based model-predictive control(EMPC). Firstly, based on the Lyapunov stability theory, two kinds of feasible regions which are with and without disturbances are defined, and the system state is detected in real time. Then, the corresponding controller is implemented for economic optimization, state driving and stability driving respectively according to the different regions which the system state belongs to. This method provides more degrees of freedom and on line optimization time for economic optimization on the promise of guaranteeing stability, therefore obtaining higher economic performance than traditional methods. The proposed EMPC designs are illustrated through a negative resistance oscillator example and their feasibility and effectiveness are evaluated through simulations.