橢柱快速高溫爐的溫度控制系統

Abstract

本論文研究之目的乃在於設計一更好的RTP高溫爐溫度控制系統，以改進控制器的溫度表現。使用者透過一簡單的輸入介面，而對整個RTP 高溫爐系統進行溫度控制。本論文中，我們將整個RTP 系統設計分成兩大部分。第一部份在去年完成，先建立熱流模型，再發展系統 判別法的步驟，以及溫度控制器的理論設計和模擬；第二部份則是我們量測機臺的特性和響應，根據量測的資料設計和製造出一套完整的控制器。我們利用Ziegler-Nichols的最佳PID參數調整法所得到的參數為依據，並輔以傳統控制法則，得到最適當的PID參數。由於系統屬於非線性的系統，我們將目標溫度分成二個區段，一是從500℃到300℃，一是從300℃到25℃。我們針對此設計出二組PID參數。另外，控制器是以8015為主體而設計的。 由於我們的實驗數據顯示本RTP 系統的特性並非是一簡單的主極點函數可表示，而事實上是二個甚至三個以上長延遲時間加低頻極點函數的綜合。依據這樣的系統，我們為其設計了一個較合適的PID控制器，原來的操作溫度200℃，誤差為±20℃，在使用了我們的控制器後，溫度控制能穩定到200℃時誤差為±2℃，400℃的誤差為±4℃，這樣的結果雖不甚理想，但已將缺點大大改進。

The objective of this thesis is to design a better temperature control of an RTP furnace to improve the temperature control rperformance given by the original controller. The user can control the temperature of the furnace through a simple input interface. In this thesis, we categorize our design work into two major parts. The first part was finished in last year. The thermal model of the furnace has been built up and the system identification steps and theoretical design of the temperature controller was proposed and simulated in last year. In the present part, we measure the characteristics and responses of the furnace and then use the measured data to design and fabricate the controller. We use both (1) Ziegler-Nichols optimun PID design to obtain the PID parameters and (2) the heuristic control rule to obtain the best choice of the PID parameters. Because there exists serious nonlinear characteristics in the temperature response of the furnace, we divede the operation temperature regions of the furnace into two parts: one is from 500℃ to 300℃ and the other one is from 300℃ to 25℃. We design two sets of the PID parameters for the two operated temperature regions. The controller is designed in 8051 based circuit. Our measured data has shown that the temperature response of the furnace can not be represented by only a simple dominant pole. Actually two or even more low frequency poles with long delayed times are needed for completely representing the characteristics of the furnace. Therefore, we choose the PID parameters for obtaining the best temperature control. Our experiments show that the temperature deviation produced by the original controller is about ±20℃ at 200℃ operating temperature. After using our controller the temperature deviations are eliminated to ±20℃ at 200℃ and ±4℃ at 400℃. Although our controller is not good enough, the performance of the furnace has been improved a lot.