Real-time temperature field reconstruction of boiler drum based on fuzzy adaptive Kalman filter and order reduction

Based on the fuzzy adaptive Kalman filter (FAKF) and an order reduction technique, a real-time on-line temperature field monitoring method for boiler drum is established. Adopting the measured temperatures of the drum outer wall, the FAKF and weighted recursive least squares algorithm (WRLSA) are used to acquire the internal heat flux and reconstruct the temperature field of a boiler drum inversely. In the above process, the aggregation method is developed to reduce the orders of the heat transfer model, by which the accurate reconstructed results can be achieved using less measurement points. In addition, using the filter residual, the process noise covariance of the Kalman filter (KF) is adjusted by fuzzy inference. Thus, the stability of the technique for temperature field reconstruction is improved. The start-up curve of a 600 MW subcritical boiler is used to verify the effectiveness of the proposed method.