Please use this identifier to cite or link to this item:
標題: 變頻式油溫冷卻機之模型參考適應預估控制
Model Reference Adaptive Predictive Control of a Variable-Frequency Oil-Cooling Machine
作者: 鄧斐仁
Teng, Fei-Jen
關鍵字: 油冷卻;oil-cooling;模型參考;預估;model reference;predictive
出版社: 電機工程學系
本論文的主要目的在發展一以數位信號處理器 (DSP) 為基礎的變頻式油溫冷卻機控制器之模型參考適應預估控制法則及其實現技術。油溫冷卻機之近似系統參數可由遞迴最小平方法則(Recursive Least-Squares Estimation Method)來求得。首先提出一利用德州儀器公司 TMS320F243 DSP 單晶片來作為運算處理單元並結合相關週邊介面發展出一低價單板高性能溫度控制器。為達到良好的溫度控制效果,本文提出二種模型參考控制策略:模型參考間接自調預估控制 (Model Reference Indirect Self-Tuning Predictive Control) 及直接自調預估控制 (Direct Self-Tuning Model Following Predictive Control)。經由實驗結果證明本論文所提的控制策略,在不同的實驗條件下,例如:設定點改變、外加固定負載及負載改變,皆能有效地在DSP-Based 控制器上得到符合規格的控制性能。

This thesis develops methodologies and techniques for model reference adaptive predictive control laws design and implementation of a DSP-based controller for a variable-frequency oil-cooling machine. The plant is well approximated by a first-order system model with time delay and its system parameters are estimated using the recursive least-squares estimation method. In order to obtain a low-cost, real-time and high-performance controller, we build a stand-alone control architecture using a powerful TMS320F243 DSP single chip with associated peripherals. To achieve satisfactory control performance specifications, the thesis proposes two model reference adaptive predictive control strategies, model reference indirect self-tuning predictive control and direct self-tuning model following predictive control. Experimental results show that both the two proposed control methods together with the DSP-based controller's prototype have been proven to work well for the machine under set-point changes, fixed heat loads and heat load changes.
Appears in Collections:電機工程學系所

Show full item record

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.