Source seeking via stochastic approximation algorithm with expanding truncations

摘要

In this paper, stochastic approximation algorithm with expanding truncation is developed to solve source seeking problem with velocity actuated and force actuated vehicles. We give necessary and sufficient conditions for the convergence of the algorithm and compared with the existing work, a larger amount of measurement noise can be considered. At each iteration, only two observations are required to estimate the gradient of the signal field function by the means of a Kiefer-Wolfowitz algorithm. The effectiveness of the proposed algorithm is illustrated through numerical simulation.