本論文提出一以數位訊號處理器為基礎之TSK 機率模糊類神經網路智慧型控制器以控制磷酸鋰鐵電池儲能系統，此電池儲能系統具有電池管理系統與雙向功率流動之三相交流-直流轉換器，本電池儲能系統並可對電網以實虛功控制策略進行併網及充電。為了要改善功率在命令變動時之暫態響應，本文因此採用TSK 機率模糊類神經網路控制器以取代傳統的比例積分控制器。本文將詳細介紹TSK 機率模糊類神經網路的架構以及線上學習法則，而所提出之智慧型電池儲能系統皆實現於以32 位元定點運算之數位訊號處理器TMS320F28035 上。另一方面，為了增強數位訊號處理器之運算效率，本論文將以組合語言撰寫所推導之控制法則。最後，將由實驗結果驗證所提出之TSK 機率模糊類神經網路控制器實現在此電池儲能系統上之控制性能。A digital signal processor (DSP)-based TSK-Type probabilistic fuzzy neural network (TSKPFNN) is proposed in this thesis to control a 4 LiFePO batterystorage system. The storage system includes 4 LiFePO battery module with battery management system (BMS) and bidirectional power flow three-phase AC-DC converter. Moreover, the designed storage system adopts active andreactive power control for grid connection. Furthermore, to improve the transient of command variation, a TSKPFNN controller is proposed to replace the traditional proportional-integral (PI) controller. The network structure and the online learning algorithms of the TSKPFNN controller are introduced in detail. In addition, all the control algorithms for the proposed battery storagesystem are realized in a 32-bit fixed point DSP, TMS320F28035, using assembly language for enhancing the calculate efficiency of the DSP. Finally, the controlperformances of the proposed TSKPFNN control system are evaluated by some experimental results.
