Hybrid differential evolution with geometric mean mutation in parameter estimation of bioreaction systems with large parameter search space

Problems of parameter estimation of nonlinear bioreaction systems are in general formulated as function optimization problems and are known to be frequently ill-conditioned and multimodal. While the optimization problems are defined on a large parameter search space, only few evolutionary algorithms are able to find a global solution to the large parameter search problem. In this study, a geometric mean mutation was embedded to hybrid differential evolution to replace a gene of the selected individual outside the assigned region. The replaced individuals were then applied to a differential mutation strategy to yield a perturbed individual. The benefit of using a large parameter search space to an inverse problem is to reduce the kinetic model complexity to yield a more compact formulation. Two inverse problems and twelve static benchmark problems with the large parameter search space are employed to illustrate the effectiveness of the proposed method.