A General Online Trajectory Planning Framework in the Case of Desired Function Unknown in Advance

Trajectory planning approaches including off-line and on-line algorithms are developed to deal with physical constraints in practical systems. However, by now the existing trajectory planning algorithms have to assume that the desired trajectory function to be planned is fully or at least partly known in advance, and it may not be true in some applications. To overcome this limitation, a general framework of on-line planning a desired trajectory under physical constraints whose function is unknown in advance is proposed. The desired trajectory is first on-line interpolated to achieve amathematical expression, and then planned under the physical constraints by the bound estimator and nonlinear filter. A heuristic critical test curve algorithm is proposed to solve the potential stability issue. A telerobotic system, where the function of slave-desired trajectory is unknown in advance, is selected as a typical case. The experimental results validate the effectiveness of the proposed planning algorithm.