本論文提出以小波模糊類神經網路智慧型控制器實現鼠籠式感應發電系統併聯市電之控制，並推導以智慧型混合控制器控制鼠籠式感應發電系統之直流鏈電壓。此系統在固定風速及變動風速的情況下，皆能準確地偵測市電角度，並提供穩定的實功率與虛功率給市電。本論文以磁場導向控制此鼠籠式感應風力發電系統，並以小波模糊類神經網路控制器來改善此系統操作在不同條件下的暫態和穩態響應。小波模糊類神經網路控制器以倒傳遞學習演算法進行線上訓練，分別透過交流轉直流轉換器控制直流鏈電壓與直流轉交流轉換器控制實功率和虛功率的輸出。論文中將詳細推導小波模糊類神經網路控制器之網路架構與線上學習法則，以及智慧型混合控制器的控制架構與穩定性分析。另一方面，亦採用PSIM軟體模擬鼠籠式感應發電系統之可行性，最後透過實驗結果來驗證控制方法之有效性。This thesis presents a wavelet fuzzy neural network (WFNN) intelligent controller to control the squirrel-cage induction generator (SCIG) system for grid-connected power application, and a hybrid intelligent controller to control the DC-link voltage of squirrel cage induction generator system.This system can detect the phase angle of the grid accurately and also provide a stable active power and reactive power to the grid at the testing conditions of the fixed speed and the variable speed of the wind. The field-oriented mechanism is implemented for the control of the SCIG system in this thesis. Moreover, the WFNN intelligent controller is proposed to improve the transient and steady-state responses of the SCIG system at different operating conditions. The on line trained WFNNs using backpropagation learning algorithm are implemented as the controllers for the DC-link voltage of the AC/DC power converter and the active power and reactive power outputs of the DC/AC power inverter. Furthermore, the network structure and the on line learning algorithm of the WFNN are introduced in detail. In addition, the control scheme and the analysis of stability of the hybrid intelligent controller are also introduced in this thesis. Additionally, some simulated results are given to verify the design of the SCIG system via PSIM. Finally, the feasibility of the proposed control scheme is verified through experimentation.
