Robust soft sensors for SBR monitoring

In wastewater treatment some process variables may be very difficult to mea sure directly because of the non-availability or excessive cost of sensors. An alternative is to use "soft sensors" that provide online estimates of t hese inaccessible variables by calculations based on auxiliary measurable v ariables. Two such sensors are proposed based on extended Kalman filtering or neural networks that could enable the monitoring of nitrate ion, ammoniu m ion and carbonaceous matter concentrations during nitrification of wastew ater.