以模糊邏輯設計及DSP實現之氣體流量控制器

Abstract

本論文提出一個以TMS320C50數位訊號處理器(DSP)實現使用系統判別和模糊邏輯原理來設計之氣體流量控制器(MFC)。在目前的流量控制器中，不論是數位式或類比式，大都是應用PID來控制，但若PID控制器的參數固定時，當有大量的氣體變化時，很容易產生超越量(Overshoot)，可能造成流量控制器閥門之損害。因此，我們應用著名的Steiglitz-MacBride法來判斷在不同壓力下、不同氣體流量設定值的流量系統受控體模型。利用此方法得到的系統參數，設計以主極點決定法為基礎的比例-積分控制器(PI-Controller)。接著引進模糊邏輯的觀念，取代PI控制。由於模糊控制可以根據受控體的狀態改變來自動調整其控制參數，因此可以結合同一壓力下、不同流量設定值的各個PI控制參數，成為一個新的流量控制器，並且避免超越量的發生。我們使用TMS320C50來實作此一模糊控制器與驗證其性能及優點。本論文將顯示令人滿意的實驗結果。

In this paper, we introduces a new mass flow controller using the technique of system identification and fuzzy logic. We realize it by the digital signal processor(DSP),TMS320C50. Currently, most digital and analog mass flow controllers use PID control technique. However, if the parameters of PID controller are fixed, the overshoot problem during a large change of mass flow could damage the valve of mass flow controller. Thus, we apply the well-known Steiglitz MacBride method for identifying the plant model under different pressure and flow desire value. The resulting system coefficients are used to design PI controllers based on the dominant pole positioning procedure. Then we employ the fuzzy logic to replace PI control. Since the parameters of fuzzy controller can be automatically tuned according to the states of system plant, it can associate with effects of each PI controller and form a new mass flow controller so that the overshoot can be avoided. We implement the fuzzy logic-based mass flow controller by a TMS320C50 and verify its performance and benefits. The experimental results are satisfactory and shown in this paper.