We have developed an algorithm that automatically executes step response tests for defining position correspondence, which is the most important tuning factor in basis weight profile control for paper machines. The algorithm is composed of two parts. One is the part that defines the slice bolts that are manipulated and their manipulation direction by using neural networks that have lateral inhibition. The other is the part that optimizes the amount of manipulation by using wavelet analysis. We show an example where step response test was executed automatically by actually using this algorithm and show the effects of this algorithm.