2007 International Symposium on Nonlinear Theory and its Applications
A new parameter adjustment approach for solving vehicle routing problem with chaotic neurodynamics
Takashi Hoshino, Takayuki Kimura, Tohru Ikeguchi,
PDF download (158KB)
We have already proposed a solving method with two simple local searches driven by chaotic neurons for vehicle routing problem (VRP). The chaotic neuron qualitatively realizes refractoriness that makes it possible to memorize past searching history with an exponential decay. Therefore, the proposed method can escape from local minima e?ectively. We con?rmed that the proposed method shows good performance for the VRP with time windows. However, in order to obtain good performance, it is inevitable to adjust an optimal parameter set. In this paper, we propose a new parameter adjusting approach for the proposed chaotic searching method. Using this approach, the performance of the proposed method is much improved without ?ne adjustment of parameters of the chaotic neurons.