Heuristic Search Algorithms for the Determination of Rate Constants and Reaction Mechanisms from Limited Concentration Data
David B. Terry, Michael G. Messina · Journal of Chemical Information and Computer Sciences · 1998
We present a chemical kinetics search algorithm that can be used to predict rate constants and reaction orders for chemical reactions from a limited set of concentration versus time data. The algorithm is based on the functional optimization of a well-defined error function. This error function is defined as the squared difference between the known set of limited concentration vs time data and the concentrations that are computed from a postulated set of rate constants and reaction orders. The best sets of rate constants and reaction orders are found by minimizing this error function. The error function contains many local minima, and therefore a heuristic search algorithm is used to find a useful local minimum in the error function. In this work we critically compare the performance of the chemical kinetics search algorithm when two different heuristic optimization schemes are used: the Genetic Algorithm and Simulated Annealing. We test the search algorithm on three types of chemical reactions: a consecutive reaction with a reversible step, a parallel reaction, and a reaction that represents the chemical fate of a pollutant in the environment.