کنترل تعادل از اجزای گیاه برای سیستم های پیل سوختی اکسید جامد با حساسیت به تشکیل کربن
Abstract: Solid oxide fuel cell systems have the potential to provide efficient, low greenhouse gas emitting power without the availability problems of both wind and solar energy. SOFC systems operate at high temperatures (600 C) in order to reduce ionic transport losses through a ceramic electrolyte. The benefits of the ceramic electrolyte include not requiring platinum based catalysts and a robustness to fuel composition. However such high temperatures create engineering challenges in construction, operation, and durability of the system as a whole. Both fuel and air must be pre-heated prior to entering the fuel cell stack. In order to ensure that carbon does not build up and degrade the system some form of fuel preprocessing is required. To move air and fuel through the system, blowers and valves must be used. Additionally during system start up, a method for pre-heating the fuel cell to within an operating range is required. All these components are tightly coupled to the time response and overall performance of the system. They also all have constraints and operating ranges, for example the fuel reformer must remain within a set temperature range or risk damage. Thus model predictive control is a natural choice to ensure that the maximum load following and overall system efficiency can be maintained without damaging components. This thesis analyzes system wide control of a solid oxide fuel cell system comprised of a tubular stack bundle, fuel reformer, air pre-heat exchanger, tailgas burner, and air blowers. Control oriented, dynamic component models have been created, allowing for estimation of temperatures and gas compositions throughout the system. The effects on system response of each component is analyzed, providing insight into realizeable response to load changes and sensitivity to noisy input parameters such as varying fuel stocks.
Keywords: Applied sciences, Solid oxide fuel cells, Carbon formation, Fuel cell control, Fuel cell modeling, Linear parameter varying, Model predictive control, Sofc, System identification