Study On Modeling And Accuracy Of Thermal Power Plant 600MW Boiler

For the design and optimization of thermal power boiler automatic control system, the establishment of high precision controlled object models and master its dynamic characteristics are developed to improve the thermal control system prerequisites. Due to thermal power boiler system with high parameters, strong coupling, multivariable characteristics, use mechanism modeling method tends to produce large systematic errors, the model of the dynamic characteristics is difficult to achieve satisfactory results.For recent years, distributed control systems and plant-level monitoring information system widely used in thermal power plants, so that the site can easily get in the field to run the historical data and real-time data. Fully exploit these data, select the field contains operational data to reflect thermal system dynamics system modeling study was a real sense of having. With the continuous development of systems theory and modeling techniques to identify and improve the application of modern technology to identify the areas of thermal system modeling has practical significance.This paper outlines the working principles and tasks drum boiler and once-through boiler and boiler system differences and focus to 600 MW supercritical once-through boiler example to explain the system features. Then, the analysis and comparison of the theory and the classical theory of modern identification to identify and demonstrate the least squares method to identify the advantages and disadvantages of PSO with specific examples. Because of the least squares method to the system input signal has more stringent requirements for the thermal model identification system has its limitations. In this paper, based on historical data using particle swarm site boiler system model identification method, and has been a high precision system model. Before model identification, first select a certain operating point excitation signal better historical data, and then the selected low-pass filtering the data preprocessing thick value processing and zero initial value, using matlab programming tools and PSO identification method for boiler superheated steam temperature system, to water and furnace pressure control system model identification systems, will be the model identification data contrast with the original curve curves. And presents qualitative and quantitative methods to verify the accuracy of the simulation model, and select the same condition in different periods of historical data to run the model accuracy for a reasonable argument.In this paper, the typical difficulties boiler thermal system to fully exploit its historical operating data inherent in the field of information and the establishment of a high precision quantitative model uses intelligent identification method, this method avoids the actual site for the system to do modeling and experimental perturbations. The work that provide powerful reference basis for the design and optimization of thermal power plant automatic control system.