基于自适应罚函数的QPSO算法的代谢通量评估

Metabolic flux estimation based on quantum-behaved particle swarm optimization with self-adaptive penalty function

针对代谢通量评估问题属于带约束的优化问题,其目标函数是一个非线性、不可微的并且存在多个局部最小点的复杂函数,提出了使用自适应罚函数的量子粒子群优化算法来解决这个问题。通过自适应罚函数的方法解决约束条件,然后使用QPSO算法最小化内部代谢通量。用此算法评估谷氨酸棒杆菌的内部代谢通量并与传统的优化算法来比较,实验结果证明了该算法的可行性和有效性。

Metabolic flux estimation corresponds to a constrained optimization problem,objective function of which is non-linear and non-differentiable and exists multiple local minima making this problem a special difficulty.This paper proposed quantum-behaved particle swarm optimization（QPSO） with self-adaptive penalty function to solve 13C-based metabolic flux estimation problem.The method transformed the stoichiometric constraints to an unconstrained one,by penalizing the constraints and building a single objective function,which in turn was minimized using QPSO algorithm for flux quantification.The proposed algorithm was applied to estimate the central metabolic fluxes of corynebacterium glutamicu and compared with conventional optimization technique.Experimental results illustrate that this algorithm is feasibility and validity.