Team:TUST China/Model

Team:TUST China 2017


The Surface Response Method in Modeling

  • Due to the failure that wet-lab experiments could not provide the subjective map despite the design and optimum methods of experiments, therefore it is not right for us to judge the optimal point merely by intuition. Though the optimal value is available, it is barely for us to tell the optimum area. In this circumstance,we utilized RSM(Response Surface Method)to analyze our formerly collected data and optimize the fundamental culture medium of the co-culture system.

    Here are what we principally concerned about when we designed our RSM modeling:
    1. First, we hope that our RSM modeling is able to target and solve out the obstacles which the PH reduced too much by the end stage of fermentation.
    2. Second, we hope that the yields of BC could be improved by the culture medium modified by RSM.

    The following introduction will introduce how we implement the RSM modeling to achieve our goal.

    First of all, our team set up the upper and lower limit value of carbon source and the quality of enhancing factor——ethanol according to the relevant literatures we formerly read and the results from pre-experiments. With these statistics,we have confirmed how should we design our experimental samples below:

    (Figure 1. experimental samples)

    By means of wet-lab experiment, we have obtained the PH range of culture medium residues at the end stage of fermentation, then we used multiple regression equation to analyze results from RSM. (Figure 2~4.) After that, we confirmed the range of optimal factor by parsing and estimating the effect of the PH from interactions through the group dynamic figure for any factors affecting acid fermentation end stage production.

    (Figure 2. RSM space diagram of Glycerol and 50% EtOH)

    (Figure 3. RSM space diagram of Glucose and 50% EtOH)

    (Figure 4. RSM space diagram of Glucose and Glycerol)

    From figure 2~4, despite regression model flatten out, it is obvious that the optimal value does exist for the analysis of various factors on the changing trend of response values.

    Software is used to analyze ridge analysis, this is the table we designed when the PH at the favorable range and yield intend to close to the optimal value:

    formula Glu(g) Gly(ml) 50%EtOH Predicting Final pH
    Target 15.11 9.53 0.05 4.36114