最优化中利用惩罚因子的罚函数法的外延，连带了共轭梯度法，由于罚因子的迅速跌代增大，加速了目标函数的收敛速度。-optimization of the use of punitive factor function in the epitaxial sustaining a conjugate gradient method, As the penalty factor to the rapid increase or on behalf of, the objective function to accelerate the convergence rate.