广义卡尔曼滤波算法在真菌发酵过程状态估计中的应用

APPLICATION OF EXTENDED SEMICONTINUAL MULTIRATE KALMAN-FILTERING TO STATE ESTIMATION OF FUNGUS FERMENTATION

  • 摘要: 真菌深层培养过程是一个相当复杂的生物化学过程.为了对过程进行优化,需建立动力学模型对发酵过程进行模拟.而在发酵过程中存在着诸如菌类变异、设备与环境变化等不确定因素所形成的随机噪声,这类噪声县确定性模型无法预估的.为了提高数学模型的模拟精度,本文采用广义卡尔曼滤波技术进行递推滤波估计.计算结果表明,利用这种滤波处理可以改善状态变量的估计精度.

     

    Abstract: The suspended culture of fungus is a quite complex biochemical process.The modelling offungus fermentation is necessary for optimizing this process. However,there are some undetermined factorsduring the fermentation process,such as variety of microorganism, changes in facilit and circumstances,Some of these would cause stochastic noises disturbing the bioprocess. In this paper an extended Kalman-fil-tering was used to improve the precision of the simulation of mathematical model.The results of calculationshowed that the precision of the estimation of state variables could be ameliorated by using this filter.

     

/

返回文章
返回