With the emergence of electric propulsive technologies in the past decades, electric vertical takeoff and landing (eVTOL) aircraft gain increasing interest, which paves the path for advanced air mobility. However, there is insufficient prior-knowledge and historical data for performance assessment of eVTOL, especially in the low-speed performance analysis, which is critical for operational safety and efficiency. To this end, this paper utilizes trajectory optimization to provide baseline reference for control design. An algorithm is proposed to generate the minimum-time trajectory of eVTOL. To guarantee computational efficiency, motion primitives are derived based on abstracted equations of motion. Thereafter, the possible types of optimal trajectories are discussed regarding mission scenarios and verified by direct collocation method. The potential application of the proposed method is showcased in the design optimization of reference model for control law. The results show that the optimized reference model reduces control resource consumption and improves transient behaviors of eVTOL aircraft.

This article proposes a desensitized optimal control (DOC) methodology to enhance the trajectory robustness of electric aircraft. First, a comprehensive multiphysics model of electric aircraft is built based on the first principle. An emphasis is put on the battery pack characteristics led by the cell unevenness. Thereafter, the optimal control technique is applied to study the influence of thermal effects and battery unevenness on flight performance. Furthermore, to alleviate the impact of cell unevenness, the crude control scheme is augmented by DOC. The proposed method generates a more robust reference trajectory for aircraft against cell unevenness. Processor-in-the-loop tests prove that the trajectory is trackable via a tracking controller. Compared to the standard trajectory optimization and nonlinear model predictive control, the proposed method significantly reduces the sensitivity of battery pack temperature regarding cell unevenness with acceptable computational effort.