本文讨论了非线性动力生化过程的参数估计(反问题),以1个包含36个参数的3阶段代谢途径为研究对象,其数学模型描述为受1组非线性代数-微分方程约束的非线性规划问题,由于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。智能优化算法由于其高效性、收敛性和鲁棒性等特点被广泛应用于非线性问题优化,于是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,建立求解参数估计的算法模型,采用人工模拟实验值,通过改变已知参数值增加试验次数减少实验误差,将参数编码成算法的1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。仿真试验表明用量子粒子群算法求解较好,该算法有效地估计了模型中的36个参数,并且成功地完成了对已知模型的预测。
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 problem is frequently ill-conditioned and multimodal,traditional(gradient-based) local optimization methods fail to arrive at satisfactory solutions.The use of several intelligent optimization algorithms has been explored.A case study considering the estimation of 36 parameters of a nonlinear biochemical dynamic model has been taken as a benchmark.Quantum-behaved Particle Swarm Optimization algorithm is able to solve the problem successfully has been showed in experiments.