[ \exists t: | q_actual(t) - \tau_planned(t) | > \sigma \cdot \textVar s \in [t-\delta,t] \left[ \frac\partial c obs\partial q(s) \right] ]
You're looking for a deep review of hdmove2! hdmove2
RRT* and PRM* provide probabilistic completeness but require dense sampling in high-dimensional configuration spaces ( Q \subset \mathbbR^n ). For ( n > 30 ), the volume of ( Q ) grows exponentially, leading to impractical sampling densities [4]. [ \exists t: | q_actual(t) - \tau_planned(t) |