Calcium Carbide Furnace Multi-step Prediction And Design, Self-tuning Control System

Calcium Carbide is extensively used in industry production as chemistry material. In conventional Calcium Carbide production process, adjusting three phase electrode lift is the key problem in Calcium Carbide furnace control system. For the three phase electrode lift has some characters as multivariate variable, strong coupling, non-linear and so on., PID controller often doesn't have good control effect in this system. So the controller based on Adaptive control algorithm were researched to be used in three phase electrode lift control system.The main work of this thesis is list as follows:1. To begin with, the principle and flow of the Calcium Carbide furnace production were introduced. Furthermore, the hardware and the software of system were selected. The Siemens PLC was selected as subordinate computer and WinCC was selected as monitoring software in this system.2. The model of three phase electrode lift was built. This model has two parts. One is Lift hydraulic unit model which was built with Step Response method. The other is electrode location- electric current model which was built with Least Square method.3. The Multi-step Predictive Self-tuning control algorithm was researched in three phase electrode lift control system. Through the simulation in MATLAB, Multi-step Predictive Self-tuning control algorithm were proved to have better control effect than the PID algorithm and the PID algorithm based on the structure of indirect self-tuning had in electrode location-electric current model. So The Multi-step Predictive Self-tuning control algorithm was ??cted as control tactic in electrode lift control system.4. The design of electrode lift control system based on Multi-step Predictive Self-tuning control algorithm was finished. The PLC software and the supervisory control picture of WinCC were finished. Moreover, the technic of connecting WinCC and MATLAB by DDE was used to design the process parameter estimator in the system. This method ravel out the WinCC's weakness in scientific computation.