A risk-aware reference trajectory resampling method for quadrotor tracking accuracy improvement

Purpose Many existing trajectory optimization algorithms use parameters like maximum velocity or acceleration to formulate constraints. Due to the ignoring of the quadrotor actual tracking capability, the generated trajectories may not be suitable for tracking control. The purpose of this paper is to design an online adjustment algorithm to improve the overall quadrotor trajectory tracking performance. Design/methodology/approach The authors propose a reference trajectory resampling layer (RTRL) to dynamically adjust the reference signals according to the current tracking status and future tracking risks. First, the authors design a risk-aware tracking monitor that uses the Frenét tracking errors and the curvature and torsion of the reference trajectory to evaluate tracking risks. Then, the authors propose an online adjusting algorithm by using the time scaling method. Findings The proposed RTRL is shown to be effective in improving the quadrotor trajectory tracking accuracy by both simulation and experiment results. Originality/value Infeasible reference trajectories may cause serious accidents for autonomous quadrotors. The results of this paper can improve the safety of autonomous quadrotor in application.