Adaptive Filtering For Thermal Process Signal And Its Application In Modeling And Control

For the booming of information process in our country's power unit, DCS (Distributed Control System), SIS (Supervisory Information System in plant level), MIS (Management Information System) have been widely used as the key part of the process control system, information monitoring system and management system. Advanced control algorithm, computing function, diagnostics and optimization have been applied in those platforms, which accelerate the control level, operation level and management level. The implementation of these algorithms needs a lot of signals from the field. So the reliability, the accuracy and the consistency of those signals plays a very important role. Therefore, it is necessary to do the data preprocessing job for those field signals. Otherwise, it is essential to have knowledge about the controlled object for theoretical study and engineering. So it is of great significance to learn a signal processing technology for the thermal process signal. In this paper, take the adaptive filter theory as a tool, and use it to eliminate the noise in thermal process signal, build the model for the unknown thermal object, estimate the delay of the thermal process object, and incorporate with the model algorithmic control.The main content and research result:Study the adaptive filtering theory which includes Least-mean-square algorithms and Recursive-least-square algorithms. According to the characteristics of the thermal process, part of the algorithm is improved for better using to the analysis of thermal processes and signal processing. For the existing noise disturbance problem in industrial process monitoring and control, implement ANC (Adaptive Noise Canceling) technology for the noise suppression and elimination to the industry process signal and control.A thermal process modeling method based on adaptive filtering technology is proposed, the modeling results is FIR (Finite Impulse Response) model. The tap weight distribution of the FIR model provides a visible modeling result. The simulation results show that this modeling method has strong robustness, and the algorithm is simple, easy to implement.Study the relevant theories of the TDE (Time-Delay Estimation), verify the adaptive time delay estimation has a very good performance for suppression the noise to the input. For the delay of the thermal process object, proposed a method of estimating delay for the process object, which is based on modeling method, and simulation result shows its perfect effect.Study the algorithm theory of the model algorithmic control, and find the relationship between the controller parameters and the prediction model during the simulation. The model algorithmic control method is improved to have a more accurate prediction models and low computational complexity. Combine the adaptive modeling method based on adaptive filtering and model algorithmic control, proposed an adaptive model control algorithm based on adaptive filter, and the simulation results show the effectiveness of this approach.