Design Of Flash Dryer Control System And Research Of Temperature Control Methods

Improving economic efficiency by energy saving and consumption reduction is the pursuit of goals of modern business. In this context, dryers are widely used in the metallurgical, chemical, pharmaceutical, environmental protection and other industries. Recently, most of the drying processes use manual control or conventional PID control methods. Although these control methods for small, single productions can meet the requirements, they are ragged when dryers and other equipments are combined into a larger system. With the dryers change to large-scale, integrated, continuous, complex, the process control of quality is forward higher requirements. Building gypsum powder drying and calcining system is a typical application of flash dryer. Therefore, this thesis designs a building gypsum powder drying and calcining control system, and focuses on discussing the control method of the main control index--dryer inlet temperature.Firstly, this thesis introduces the history, characteristics and basic structural of flash dryer, and describes the principles of drying gypsum calcining process. Secondly, according to the principle of rotary flash drying, process design, equipments, and the characteristics of gypsum drying material, this thesis accomplishes the design of drying and calcining control system, including configuration of machine hardware and software programming. Finally, the PLC programs of the switch process control and analog process control are designed, and the designs of the WinCC process picture, configuration alarm recording, trend curve, print statements are also accomplished.Based on the conventional PID control method, this thesis improves the design of PLC control system, and uses the BP neural network for dryer inlet temperature control system modeling. Because flash dryer inlet temperature system is a multivariable, nonlinear, slowly time-varying time delay system, we can hardly build an accurate model, which is necessary for the traditional control methods. On the basis of summarizing the existing control methods, this thesis proposes a neural network and Smith predictive control method of the dryer temperature, and provides controller design and simulation.The Smith neural network control method in this thesis has a better effect of the dryer entrance temperature control. By learning the advantages of the neural network and the traditional control method, and using the accurate modeling ability, adaptive ability, self-learning ability of neural networks and overcoming system delay ability of Smith control, it combines the neural network with the traditional control method effectively. It has theoretical significance to control the process of large time delay which is common for industry.