把 Chebyshev多项式和偏最小二乘法有机地结合起来 ,提出一种基于 Chebyshev多项式的自适应偏最小二乘回归建模算法 .它自动建立一个预测误差较小的基于Chebyshev多项式的多元多项式模型 ,它的预测效果优于传统多元多项式模型 .本文算法是偏最小乘法的推广 ,也是偏最小二乘法在非线性领域里的应用 Combining the Chebyshev polynomials and partial least squares, an adaptive partial least squares regression algorithm based on Chebyshev polynomials is proposed, which automatically establishes a multivariate polynomial model based on the Chebyshev polynomial with small prediction error. The prediction effect is better than the traditional multivariate polynomial model.The algorithm in this paper is the promotion of partial least-squares method and also the application of partial least-squares method in the field of non-linearity