積分型T-S模糊控制設計應用於非線性切換式電源轉換器

本論文提出一個簡單化的積分型T-S 模糊控制器，作為具有高度非線性特質的直流至直流升壓型切換式電源轉換器及交流至直流隔離型主動式高功率因素校正轉換器的輸出電壓調節。為了確保優異的穩定狀態，我們導入一個額外的積分誤差訊號到動態式子裡。經由轉移座標到要調節的工作點上，既可將穩定化模型確立。然後，建立此積分型的T-S 模糊模型，做為處理非線性項之用。控制增益及系統穩定度可同時經Lyapunov理論推導，經由求解線性矩陣不等式後可得控制器增益值。令人驚奇的性質是這些增益值在每一條控制法則上具有一致性，這結果大為簡化了控制器型式，成為一線性狀態回授型控制器。此控制器已經由類比電路實現。甚且應用於非線性切換式電源轉換器，其數值模擬及實驗結果皆顯示良好的性能，且具備設計所需之滿足的衰減比率。另一方面，當作輸出調節的標準化問題探討，如何針對非線性切換式電源轉換器設計有效的無電流感測控制器之議題也被展現。基於積分型T-S 模糊化模型，兩種控制方法被提出及深入地研究:(1)使用虛擬想要的參數(VDVs)之輸出調節，與(2)經由座標轉換的穩定化方法。值得強調的是，拜積分型控制架構之賜，這兩種方法引導出簡單的控制器。考量運用的形式，經由包含觀察器結合的控制器以感測迴圈電流，無感測型的電源轉換器被完成了。由於模糊隸屬函數滿足類似Lipschitz的性質，經由求解線性矩陣不等式，控制器增益及觀測器增益可分開獲得。進一步地，這兩種方法的內在關係與比較被揭露了。在緊密的探究後，於此宣告這兩種方法實際上具有相同的型式。虛擬想要的參數之方法能被運用在輸出調節與輸出追蹤控制上，座標轉換法則可被使用於在一固定點上做穩定化。最後，這兩種無電流感測的控制器被實現，並應用於降壓型的切換式電源轉換器，以Matlab/Simulink 的模式做數值模擬，並經由數位訊號處理器的方式做實作實驗。本論文已發展出無電流感測形式的控制策略，並且顯示其可行性及滿意的性能。

關鍵字

Lyapunov理論；切換式電源轉換器； T-S模糊控制器；無電流感測控制器. ；線性矩陣不等式

並列摘要

A simple integral Takagi-Sugeno (T-S) fuzzy controller is proposed to regulate the output voltages for DC-DC boost switching power converters and AC-DC isolated active high power factor correction (AHPFC) converters, which have highly nonlinear characteristics. To ensure nice steady-state, an extra integral error signal is added to the dynamics of the converters. By translating coordinate to the regulated points, the stabilization model is determined. The integral T-S fuzzy model is established for dealing with the nonlinear terms. Both control gains and system stability are inferred by Lyapunov theorem. The controller gains can be obtained by solving linear matrix inequalities (LMIs). A surprised property is that the controller gains are identical for every fuzzy control rule. This result greatly simplifies the controller to be a linear state feedback one. It has been realized by using analog circuits. Furthermore, the numerical simulations and experimental results display the nice performance with satisfying the designed decay rate. On the other hand, as a standard problem of output regulation, the issue that how to design efficient sensorless controller is also presented for nonlinear converters. Based on the integral T-S fuzzy model, two control approaches: (i) Output regulation using virtual-desired variables (VDVs); and (ii) Stabilization approach via coordinate translation are proposed and studied in depth. It is worthwhile that, thanks to the integral control structure, these two approaches lead to simple controllers. For considering the sake of applications, sensorless converters are complemented by including a state observer in the controllers to estimate the choke currents. Since the fuzzy membership functions satisfy a Lipschitz-like property, the controller gains and observer gains can be separately obtained by solving LMIs. Moreover, the comparisons and the intrinsic relations between these two approaches are explored. A closer investigation declares that both approaches are actually with the same form. The VDVs method can be applied to the output regulation and the output tracking control. The coordinate translation approach is used for the stabilization at a fixed point. Finally, both current-sensorless controllers applied on buck converters are implemented by Matlab/Simulink models for numerical simulations and by digital signal processors for experiments. It has been developed that the proposed sensorless control schemes are feasible with the satisfactory performances.

並列關鍵字

linear matrix inequalities (LMIs) ； switching power converters ； Lyapunov theorem ； Takagi-Sugeno (T-S) fuzzy controller ； current-sensorless controllers.

參考文獻

[18] E. C. Chang et al., “Digital integral action in variable structure control with fuzzy

[29] C. F. Chuang et al., “T-S fuzzy model based H∞ finite-time synchronization design

[42] C. S. Ting, “Observer-based adaptive fuzzy control of time-delay chaotic systems,”

[56] K.-Y. Lian and C.-W. Hong, “Simple integral fuzzy control for converters with highly

stable design of PI control for DC-DC converters with an RHS zero,” IEEE Trans.

國際替代計量

積分型T-S模糊控制設計應用於非線性切換式電源轉換器

全文下載

主題瀏覽