| Summary: | The boiler combustion process contains complex physicochemical changes, which is a nonlinear time-varying industrial process with strong interference and multivariate strong coupling. For this kind of boiler combustion process with typical nonlinear characteristics and complex mechanism, it is difficult to establish an accurate mechanism model by conventional modeling methods, so it is difficult to meet the new requirements for the optimal control of the boiler. The large amount of data accumulated in the operation process contains rich system information, and the data-driven modeling and control method provides an effective way for the operation optimization of the unit. Data dynamic characterization and control technology is an important means of data mining, coal-fired boiler data has obvious temporal and drift characteristics, for the current data tracking and supervision algorithms mostly lack of dynamics, real-time and stability issues, design an adaptive clustering model based on the improved growth of neural gas model (GNG), the establishment of nodes based on the probability, the range of the search, the average distance of node A node generation and deletion mechanism based on probability, range search and average distance of nodes is established to realize real-time monitoring of drift data. Finally, the experiments are carried out by analyzing the dynamic data of coal-fired boiler, and the experimental results show that the model and algorithm have stronger real-time tracking ability for dynamic drift data, and can accurately and effectively monitor and control the dynamic data of coal-fired boiler.
|
|---|