In a carbonator tower

the change of CO2 concentration can be reflected by the alteration of some other parameters

such as temperature

pressure

and gas flow in the tower. Based on this consideration

an estimator of CO2 concentration in the tower was formed

which gives the CO2 concentration by calculating the values of temperature

pressure

and gas flow with an algorithm. Considering the time varying environment of the tower

the estimator was designed with self- adaptive attribute

so that it can work effectively under varying environment. By using the estimator

the value of CO2 concentration in the tower can be predicted online. It not only provides useful information for control

but also saves the cost for measuring CO2 concentration.