Inverse Control for the Coordination System of Supercritical Power Unit Based on Dynamic Fuzzy Neural Network Modeling

In order to improve the coordinated control quality of a supercritical power unit, a coordinated system inverse control scheme is proposed. First, the inverse models for the unit load and main steam pressure are established offline with dynamic fuzzy neural network (DFNN) using historical sample data got from the full-scope simulator of a 600-MW supercritical power unit. Then the real-time DFNN inverse control scheme which combines offline training and online model correction is designed and programmed with MATLAB. Finally, control tests are carried out through real-time two-way data exchange with the simulator. The simulation results show that the load response speed under large variable load conditions is greatly improved with the DFNN inverse control scheme compared to the original PID control, which is benefit for the supercritical unit in giving full play to its in-depth peak load and frequency regulation capabilities.