Sequential Decomposition and Dynamic Activation for UAV Motion Planning Under STL Specifications

In this article, we propose a formal control frame-work for UAV motion planning problem encoded by signal temporal logic. The motion planning tasks in the article are divided into obstacle avoidance and destination-reaching tasks. For obstacle avoidance tasks, high-order control barrier functions are leveraged to ensure the UAV consistently avoids obstacles. For destination reaching tasks, a sequential decomposition strategy is proposed to regulate the rate of the corresponding predicate function, ensuring it always remains within a specified range and the arrival at the destination. Meanwhile, a dynamic activation mechanism is presented to reduce computational consumption and render the approach less conservative. The effectiveness of the proposed control framework is illustrated by the simulation of a UAV surveillance task that includes obstacle avoidance and destination reaching.