Prescribed performance incremental control for morphing vehicles with soft actor-critic based intelligent parameter tuning

To achieve high-performance attitude tracking control of morphing vehicles under complex flight conditions, this paper proposes a reinforcement learning based prescribed performance incremental control scheme. First, a novel prescribed performance incremental control (PPIC) algorithm based on a preset error trajectory is proposed. Different from the conventional prescribed performance control (PPC) that performs nonlinear transformation on the actual tracking error, the proposed algorithm applies transformation to the tracking error of the preset error trajectory, and steers the tracking error to achieve practical prescribed-time prescribed performance convergence along the preset error trajectory under arbitrary initial conditions, while eliminating excessive overshoot. The embedded incremental control technique significantly reduces the algorithm’s dependence on an accurate system dynamic model. This algorithm can be further extended by introducing a trajectory updating mechanism to avoid the singularity problem inherent in traditional PPC. Then, a soft actor-critic based intelligent parameter tuning (SACIPT) algorithm is proposed. As the upper-level regulator, the SACIPT algorithm tunes the error transformation coefficients and other core parameters of the low-level PPIC algorithm to suppress control signal chattering and further improve closed-loop system performance. Finally, the effectiveness and superiority of the proposed scheme are verified through numerical simulations.