以模式預測控制理論應用於固態氧化物燃料電池安全模式下之負載跟隨控制

Abstract

固態氧化物燃料電池(SOFC)藉由在陽極與陰極輸入適量的反應氣體，經過電化學反 應產生電流來供應用電。為了滿足供電目的，燃料電池輸出的電流必須隨著需求改變，因此必須建立負載跟隨系統，藉由調整可變負載滿足低負載與高負載不同情況時的電流需求。在負載變動的過程中，燃料電池的燃料使用率以及電池溫度會隨著電流改變而變化，進而使系統不穩定，導致電池的壽命減少以及結構破壞。 本文旨在發展固態氧化物燃料電池於負載變動時，在安全模式下實現穩定操控。固態氧化物燃料電池由於利用電化學反應發電，系統存在化學反應的延遲現象。為此，在陽極端，為了避免匱油現象所產生的電池結構氧化，控制氫氣流率並且採取燃料使用率控制；而在陰極端，控制空氣流率並採用電池溫度控制以避免熱應力導致的熱破壞，三種控制策略，包含：固定空氣過量比、變動空氣過量比以及理想的恆溫控制分別實現並比較分析。控制器設計均採用模式預測控制，其預測特性相當適合解決高度延遲系統的問題，藉由模式預測控制設計控制器，希望能利用其特性達到良好控制性能。在系統的建立上，使用Matlab/Simulink來建構完整系統，並做開迴路及閉迴路陽極與陰極控制系統驗證模擬。藉由在陽極與陰極的適當控制，實現固態氧化物燃料電池在負載變動的情況下穩定而且安全的操作。

The study investigates integrated Solid Oxide Fuel Cell (SOFC) load tracking system, which adjusts variable load to coordinate different current demand for electricity requirement. Under load tracking, Fuel utilization (Uf) and stack temperature are changed which may make system unstable and decrease stack life. The thesis aims to develop control strategies for realizing steady operation of SOFC in safety situation under load tracking. In order to avoid stack oxidation by fuel starvation, constant fuel utilization control of hydrogen flow rate is implemented in anode; in cathode, air flow rate is regulated for keeping the stack temperature and preventing from thermal destruction caused by thermal stress. Three control strategies in the cathode, including fixed air excess ratio control, varied air excess ratio control and ideal isothermal control, combining the constant fuel utilization control in anode are realized and compared. Model Predictive Control (MPC) generally used in chemical engineering has chemical delay characteristics and is suitable for SOFC control. MPC is adopted to design all controllers to predict the system feature and achieve good control performance. Matlab/Simulink is used to realize the dynamic simulation of the integrated SOFC system and the open-loop and closed-loop control characteristics in anode and cathode. As a result, with the proposed control in anode and cathode, the integrated SOFC system can be operated in safety under load tracking.