Joke Logo

 

 

 

Fuzzy Logic Control Module at Work

The below figure shows a test run with water in the fermentation tank and the use of N2 and O2 to demonstrate the function of the fuzzy logic control module. Imagine a fermentation where the level of dissolved oxygen is crucial for your cells to survive. If for some reason the oxygen supply should be limited or blocked, the fuzzy logic control module will decrease the substrate feed rate in small steps until it reaches zero. Once the desired dissolved oxygen level is restored, the module will increase the feed rate again.

 

Below real data from a Pichia pastoris fermentation is shown. The substrate is methanol and as the culture is adjusted to live on this substrate, the dissolved oxygen level fluctuates. To avoid cell death, the module checks the dissolved oxygen level, the rate of change of the dissolved oxygen level, and the current feed rate to see whether more methanol safely can be added or the culture should be given time to use up residual methanol in the fermentation medium.

 

The below figure shows a test run with water in the fermentation tank. O2 and pH are held constant to demonstrate the function of the fuzzy logic control module in co-opperation with the neural network software sensor. First the desired substrate concentration in the tank is set to 10 (nM, g/L or whatever). In this test, the trained neural network predicts that the concentration is around 25-30, thus the fuzzy logic control module slowly decreases the feed rate. As a second step, the desired concentration is set to 100 (green line). The sensor still outputs 25-30, thus the fuzzy logic control module quickly increases the feed rate.

 

Return to the BioStat homepage

Last modified 06 September 2001 22:13