A DNA based genetic algorithm for parameter estimation in the hydrogenation reaction

Inspired by the mechanism of the biological DNA, a DNA based genetic algorithm (DNA-GA) is proposed to determine the kinetic parameters for the hydrogenation reaction. The considered chemical process contains five reactions and 25 unknown parameters. The DNA-GA uses the DNA encoding method to represent the potential parameters and genetic operators inspired from the biological DNA are designed to find the global optimum. The study on the performance for typical benchmark functions shows that the DNA-GA outperforms the other two methods in both convergence speed and accuracy. Based on the operating data gathered from an industrial hydrogenation unit, 25 parameters are obtained by the DNA-GA and the kinetic model for the hydrogenation reaction is established. To verify the validity of the established model, another four groups of data are used to test the established model and two previously reported models. The comparison results show that the sum of square relative errors of the model obtained by the DNA-GA is the least of the test models, and its prediction is in good agreement with the practical operating data.