The Tube Furnace Redundant Adjusting Decision Support System

Tube furnace is one of the most common devices in petrochemical industry, which consumes a large amount of energy and also vents much exhaust smoke annually. In recent years, the government attaches more importance to energy-saving and emission, so the study on tube furnaces'production process on-line monitoring and optimization control has also become more and more important.Tube furnace working process is multi-variable, time-varying and non-linear, It is difficult for ordinary PID control algorithm to achieve good control effect. Therefore, according to the existing technologies and the field situation of industry, we designed the overall planning of this system. This system consists of temperature, smoke acquisition subsystem, two adjustment furnace advice subsystems based on experience and model respectively.Temperature and Smoke acquisition subsystem consists of the MCU system, temperature sensors and smoke analyzers. MCU system is the core of the acquisition subsystem, which communicates with the temperature sensors through the RS-485 bus. Smoke analyzer's output signal is 0-5v voltage. Microcontroller collects smoke signals through the AD signal acquisition module. Finally, MCU system sends the collected data to the PC through the RS-232 bus.As tube furnace working process is very complex and many factors affect the thermal efficiency, this paper studied these factors and the impact on tube furnace, as well as the engineers'practical experience, and designed the adjustment furnace advice subsystems based on experience. We chosen Visual Basic6.0 and Access as development tools, and designed a whole interface of tube furnace decision support system, and built the frame of the subsystem.According to the demands of the tube furnace control system, this paper designed the adjustment furnace advice subsystem based on model, which regarded GPC algorithm as the core algorithm. The front interface is developed by Visual Basic and the data is calculated by Matlab.The data exchanging between the front interface and internal processes are developed by mixed programming techniques with Visual Basic and Matlab.Based on the system trait, we set up the switching condition between the two subsystems, so that the adjustment furnace advice subsystems based on model can be applied to practice after improvement. We use the oxygen density fluctuation in flue smoke to determine which subsystem should be adopted. If the fluctuation range of oxygen content is small, the sub-system based on model will be adopted, on the contrary, the sub-system base on experience will be adopted.