Model Predictive Control Algorithm Based on Signal Temporal Logic with Disturbances Guaranteeing Feasibility

Abstract

In this research, we deal with model predictive control problem with disturbances based on Signal Temporal Logic specifications. In this paper, we theoretically analyze the feasibility of model predictive control problems and derive sufficient conditions for recursive feasibility. The proposed algorithm has the advantage that recursive feasibility can be determined in advance (by initial conditions). Finally, we show the effectiveness of the proposed method by confirming that the theoretical and numerical results agree with each other under generic vehicle driving scenarios, using self-driving cars.