This paper presents an evolutionary computation framework that uses genetic algorithm and particle swarm optimization to infer a threshold Boolean network of the Tryptophanase operon. A unique feature of this network is that it exhibits bistability, converging to two fixed points under certain conditions. We proposed a fitness function to achieve this, ensuring the network showed the desired dynamics, particularly the bistability property. Additionally, we explored and analyzed the results obtained from 500 simulations conducted by each algorithm. The genetic algorithm could infer 23 different networks with perfect scores, but particle swarm optimization could not infer any. The results showed that, in general, the genetic algorithm could explore the search space more effectively, obtaining networks with more edges than particle swarm optimization, thus allowing it to find networks satisfying the biological restriction of the model inferred.