| In order to improve the economic effectiveness of power plant's operation, bring the rapid development of power plant information, and further taking into account the thermal detection signal control of high-quality requirements. So the pretreatment of thermal signals become a necessary and understand of thermal processes become an indispensable component. Therefore, we need to study effective processing and analysis methods for the thermal signals and objects.Adaptive filter has good performance in an unknown environment and to track the input statistics change over time, through constantly adjust the tap weights to adapt to changes in the statistical properties of the signals, reaching the capacity of adaptive fitting signals. This Paper has studied by noise cancellation and recognition of thermal signals with different adaptive filter algorithms. And has effective control of the objects with predictive control principle.This paper mainly has discussed the application of thermal object recognition and noise elimination based on the LMS (Least Mean Squares) adaptive filter algorithm, and to improve the algorithm. Combined with recognition of the impulse response sequence, design MAC-P (Model Algorithmic Control-Proportional) cascade predictive control system with adaptive predictive control principle, through the LMS adaptive filter-line real-time identification of charged Predictive Modeling of objects, combined with improved adaptive MAC-line amendment to the controller's parameters. Simulation results show that: adaptive filter based on the recognition of such results can be well adapted superheated steam temperature system of power plants of large inertia, large delay and time-varying characteristics. The implementation of its control strategy is simple and small overshoot, control is better than PID-P cascade control; and lowering the load in the experiment, the object parameters can effectively adapt to change, to ensure better regulation performance.
|
PDF Full Text Request