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

本研究性迴归类神经网路于梯度磁场的系统参调整，这种方法不要重覆量测与调整，因所时间非常短。