This study describes the application of the feedback linearization neural networks, known from neural network computing, to the problem of gradient preemphasis.

• 本研究应用线性迴归类神经网路于梯度磁场的系统参数调整，这种方法不需要重覆量测与调整，因此所需时间非常短。