基于微分代数方法的生物除磷模型参数结构可识别性研究

Structural identifiability of parameters of a biological phosphorus removal model determined by applying differential algebra

通过将微分代数方法应用于异养菌生长Monod模型和生物除磷亚模型参数结构可识别性研究,探讨了该方法的适用性.结果表明,本研究提出的算法计算过程清晰、系统化,可以实现计算机编程自动运算;在文献中方法很难进行生物除磷亚模型参数结构可识别性研究时,采用本算法也能得出结果.本研究为活性污泥模型参数结构可识别性研究提供新的方法,并对除磷过程参数估计优化实验的设计有重要的指导意义.

A new method was developed to analyze the structural identifiability of parameters of activated sludge models. The structural identifiability of parameters of the Monod model, which describes heterotrophic organism growth, and a biological phosphorus removal sub-model were studied using differential algebra. The algorithm proposed in this paper is clear and systematic, and therefore was able to he edited by a computer. Especially, this method was applicable to the case of the biological phosphorus removal sub-model, which is difficult for other methods to describe. The results could be used to guide the optimal experiment design for parameter estimation of the phosphorous removal process.