尋找現代智慧型渦輪車輛的實際喘振邊界

Abstract

發動機裝置渦輪增壓器需要知道所引起的喘振(surge)現象的潛在損害，本文的重點是在不需要真正執行複雜的汽車實驗提出了兩個新的智慧推論，找到一個渦輪增壓汽車真正的喘振邊界(surge boundary)。渦輪增壓器喘振的發生對火花點火發動機裝置渦輪增壓器系統更具有破壞性的瞬態響應，本文首次採用非線性微分方程描述柔性管道系統中的一個渦輪增壓汽車，首先透過最小方差擬合算法的五階多項式(five order polynomial via the least-squared fit algorithm)，接下來以智慧推論方法，即利用四個物理模糊集合(Fuzzy set)，即節氣門大小、壓縮比、發動機排氣量和總有效管長度輸入到智慧推論演算系統，以取代傳統的三階多項式，來表示製造商渦輪增壓器特性圖之渦輪負荷曲線，經過這兩種演算法所得到的喘振邊界兩者相當的接近，這可精確預測渦輪增壓器之渦輪特性圖及其實際喘振邊界，在不進行任何實際的實驗情況下，透過利用建構方程式之釋放節氣門角度關鍵參數，從實際渦輪參數替代真實的物理參數帶入建構之方程式，渦輪增壓器特性圖上的喘振邊界可以準確地被預測。 為了減少系統中不希望的喘振現象和驗證本文所提出預測喘振邊界方法，在本文提出了一個新的智能控制器防止所述關鍵參數被迅速增加，同時由從發動機的空氣壓力驅動的傳統旁通閥相比，這種新的主動智能控制技術可以得到一個更少喘振現象的平滑響應。實驗使用的是SAAB B204L M95渦輪增壓發動機系統進行，實驗結果已經獲得證明，新的智能控制技術確實可以顯著的減少渦輪增壓引擎喘振的發生。

Engine with turbo charger needs to be aware of the potential damage caused by surge phenomena. The focus of this paper is to propose two new intelligent reasoning to find the real surge boundary for a turbo-charged automobile without performing real complicated automobile experiment. Compressor surge is expected to occur as a destructive transient response of a spark ignition engine with turbo charger system. A nonlinear differential equation is first adopted in this paper to describe the flexible piping system in a turbo charged automobile. In order a five order polynomial via the least-squared fit algorithm was utilized along with intelligent reasoning methods, namely the four following physical fuzzy sets - throttle size, compression ratio, engine displacement and total effective length of input. Inference calculus was also proposed to replace the conventional third order polynomial to represent the compressor load curves in the manufacture’s compressor map diagram. These algorithms were adapted to allow the prediction the surge phenomenon to be in a more accurate manner from the manufacturer’s compressor map diagram. By substituting the real physical parameters from a real turbo charged automobile into the modeling equation, the surge boundary on the compressor map diagram can be accurately predicted without performing any real experiment, however the modeling equation increases as acceleration is released. To reduce the unwanted surge phenomenon in the system, a new intelligent controller was used to preventing the crucial parameter from being rapidly increased. In comparison with the traditional bypass valve driven by the air pressure from engine, this new active intelligent control technique can yield a smoother response with less surge phenomenon. The real experiment was conducted using a SAAB B204L M95 engine with turbo charger system. Significant results were obtained to show that the new intelligent control technique is capable of minimizing the occurrence of compressor surge within a turbocharger.