均匀实验设计在化学反应动力学建模中的应用研究I.非线性代数动力学模型

An Application Study of Uniform Experimental Design in Modeling Chemical Reaction Kinetics I.Nonlinear Algebraic Kinetics Model

化学反应动力学模型能为工业反应器选型、设计、控制和优化提供重要指导。通常建立动力学模型时实验与计算工作量大、误差大,特别是高压、高温、真空或固相反应等苛刻条件下操作时取样难度高、误差大,导致模型精度低。本文以串联反应为对象,首次提出用均匀实验设计方法配置实验点,仅通过6组实验数据便获得了全部模型参数,参数估计值与真值相对误差最大为4.35%。利用统计学方法验证了模型的准确性、可靠性和显著性。不同随机误差标准差(0.001,0.01,0.1)下参数估值基本一致。不同水平数(6,7,9,12)下参数估计值与真值的相对误差可控制在6%以内。拟水平均匀实验设计的运用在保证参数估计精度的同时,减少了因素水平数。

Chemical reaction kinetics model plays a crucial role in industrial reactor selection,design,control and optimization.However,under harsh conditions such as high pressure,high temperature or solid-state reaction,the sampling is hard and the error is large,leading to low accuracy of the model.In this paper,a uniform experimental design method was proposed for the first time,taking the series reaction as the object.All the model parameters were obtained only with 6 sets of experimental data,and the maximum relative error between the estimated value and the true value was 4.35%.The accuracy,reliability and significance of kinetics model were verified by statistical methods.The estimated parameter values under different random error standard deviations(0.001,0.01,0.1)were basically the same.Under different levels(6,7,9,12),the relative errors between the estimated parameter value and the true value of the parameter were within 6%.Quasi-level experimental design could reduce level numbers without compromising estimation precision.