代谢途径智能优化的研究

摘要

研究非线性动力学生化代谢途径的参数估计(反问题),对α-蒎烯热异构化动力学模型进行研究,针对非线性代数-微分方程约束的非线性规划问题,频繁的病态和多峰值,传统的算法(如梯度算法)并不能得到满意的解.提出利用智能优化算法求解代谢途径的参数估计,利用算法的非线性逼近能力,将参数编码成算法的一组解向量,以实验值和预测值的误差平方加权的和为目标优化函数.仿真结果表明用多样性指导量子粒子群(Diversity-Guided Quantum-behaved Particle Swarm Optimization,DGQPSO)算法求解较好,算法运用自适应机制来控制算法的收缩-发散过程,使算法在一定种群多样性下保证收敛精度,有效地估计模型中的参数,证明算法有较好的求解性能.%The parameter estimation (inverse problem) of nonlinear dynamic biochemical pathways which has been stated as a nonlinear programming problem subject to nonlinear differential-algebraic constrains has been discussed. The model of mechanism for thermal isomerization of ot-pinene has been taken as a benchmark. The problem is frequently ill-conditioned and multimodal, and traditional local optimization methods (gradient-based) fail to arrive at satisfactory solutions. Several intelligent optimization algorithms have been explored to the problem. We used a weighted distance measure between experimental and predicted values as the objective function. The model for solving the inverse problem has been created. The Diversity-Guided Quantum-behaved Particle Swarm Optimization exerted an adaptive method to control the attraction-repulsion phases of the algorithm, which ensured the convergence accuracy in a diverse group. The global search capability of the DGQPSO algorithm is enhanced and the inverse problem is successfully solved as showed by comparative experiments.