摘要
In recent years, the cost of engines has become increasingly important to engine manufacturers, who are consistently faced with major problems on how to reduce cost to a minimum. Cost has become a decisive factor for aircraft design. To control the continual rapid increased cost, engine cost prediction is indispensable early in the design phase. But the cost data of an aircraft engine is small; we introduce the Robust Partial Least Squares Method in solving this problem, and reducing or removing the effect of outlying data points, which is different from the Classical PLS. We use the MATLAB software doing several simulations; results and analysis of a real turbofan engine data set show the effectiveness and robustness of the Robust PLS method. The Robust PLS method can effectively be used to estimate Turbofan Engine cost with reasonable accuracy.
In recent years, the cost of engines has become increasingly important to engine manufacturers, who are consistently faced with major problems on how to reduce cost to a minimum. Cost has become a decisive factor for aircraft design. To control the continual rapid increased cost, engine cost prediction is indispensable early in the design phase. But the cost data of an aircraft engine is small; we introduce the Robust Partial Least Squares Method in solving this problem, and reducing or removing the effect of outlying data points, which is different from the Classical PLS. We use the MATLAB software doing several simulations; results and analysis of a real turbofan engine data set show the effectiveness and robustness of the Robust PLS method. The Robust PLS method can effectively be used to estimate Turbofan Engine cost with reasonable accuracy.
