In this paper, we show how optimization methods can be used efficiently to determine theparameters of an oscillatory model of handwriting. Because these methods have to be usedin real-time applications, this involves that the optimization problems must be rapidelysolved. Hence, we developed an original heuristic algorithm, named FHA. This code wasvalidated by comparing it (accuracy/CPU-times) with a multistart method based on TrustRegion Reflective algorithm.