A Dynamic Method to Estimate Source Emission Rate and Predict Contaminant Concentrations

It is very important to develop a method to predict contaminant concentrations in an enclosed space. The key technology is source emission rate estimation and dynamic concentration prediction. This paper presented a new method to estimate source emission rate and predict contaminant concentrations dynamically. A variable-structural contaminant concentration model was built, and then the extended Kalman filter was used to estimate the contaminant source emission rate and predict the concentration based on sensor data. The model of source emission rate could be gotten by applying the least square method to the filtering data of source emission rate. Simulations were done to demonstrate the performance of algorithm. The performance of parameter estimation and state prediction could be improved by using this method, and then the accuracy and speed to predict the air quality trend could also be improved.