The Research Of Double Closed-Loop Control Of The MoSi₂ Furnace

The research seeks to explore novel techniques for more accurate and refined measurement and control of high-temperature furnaces which is a key element of numerous ore and mineral extraction processes. The control of the temperature is a key factor in influencing the viscosity of the smelt and is a key determinant of the final quality of the product. At times, a temperature difference of one Celsius can result in a change of viscosity of about 10%, which is quite significant. Thus, the importance of precise and close control of the temperature cannot be overemphasized. This report focuses on the temperature control of molybdenum disilicide (MoSi2) furnaces based on an adaptive fuzzy logic PID algorithm.The MoSi2 furnace is characterized by long time delay, non-linearity and time-variability. Approaches based purely on the traditional PID (proportional-integral and differential) controller or even the more modern fuzzy logic have all but proved inadequate. The time delay and time-variability inherent in the process present greater challenges which have proved a hard nut to crack so far.Currently, the temperature control systems employed by MoSi2 furnaces have much room for improvement in terms of accuracy and response time. This leads to substantial loss of resources, poor yield and eventually impacts the bottom line of the company. Each batch of mineral that needs to be processed requires substantial power for heating and if the temperature control is not fine enough, we may end up with a reject batch or, at the very least, obtain a poorer grade product. Therefore, temperature control of MoSi2 furnaces clearly requires better technology which can guarantee better yield and be more environmentally friendly.In the first part of this thesis, it is described the basic operation of MoSi2 furnaces and the current state of the art. It is briefly talked about the temperature and viscosity measurements. The paper then analyzes the implication of the substantial time delay. Afterwards, it is also reviewed and compared several approaches such as PID, Smith Predictive Control, Fuzzy Logic and PID with Adaptive Fuzzy Logic.In the second part, the thesis goes deeper into the theory by performing a literature review and takes a closer look at what approaches have been attempted so far. The issues and challenges pertaining specifically to the MoSi2 furnaces are highlighted. The MoSi2 furnace can be modeled as a dual closed loop process with one loop for the current and one loop for the temperature. This research then comes up with an adaptive dual closed loop fuzzy logic PID controller model. The inner loop is used to control the current with a PI controller while the outer loop is based on threshold or the adaptive fuzzy logic PID controller. This research shows that the latter adaptive approach can yield better results than the simple threshold or a pure PID technique. When properly set up, the adaptive fuzzy logic has the ability to track the changing conditions of the furnace over time and can a prior give better results.Finally, through a series of simulation, the research goes on to show that the precept is in fact correct. It is compared the result of the novel adaptive fuzzy logic PID controller with the traditional PID. It is also introduced power cut, interference and time-variability of the furnace characteristics in our simulation model to mimic the real-world as closely as possible. The results suggest that the adaptive technique can lead to closer and more accurate control of the MoSi2 furnace temperature and hence the viscometer. In addition, the adaptive fuzzy logic PID controller seems to be more robust under time-varying conditions and interference.