This paper presents a multivalued Hopfield-type neural network as a method for solving combinatorial optimization problems
with a formulation free of fine-tuning parameters.
As benchmark of the performance of the network we have used N-Queen problems. Computer simulations confirm that this network
obtains good results when is compared with other neural networks.
It is shown also that different dynamics are easily formulated for the network leading to obtain more sophisticated algorithms
with better performance.