Frequency response function technique and partial coherence function technique are two typical approaches in noise source identification of multiple-input/single-output systems. The criterion which determine the applicability of the techniques for noise source identification is the degree of coherence between the measured inputs. When the measured inputs to a multiple input system are not correlated each other, contribution of each input to the output can be obtained with no difficulty by using the frequency response function technique. For the second approach which is used in case where the inputs are correlated each other, correct decision of the priority of the input records is a key factor since the priority determines the way how the results are interpreted. This text presents a method for the determination of the priority among the inputs in applying the partial coherence function technique for the source identification problems.
The basic idea is that one signal of any two correlated signals causes the other signal and that this causality can be identified by observing the impulse response functions estimated in the negative time region. The method was applied to a simulated two-inputs/one-output system where the inputs were correlated each other and the feasibility for practical applications has been demonstrated through a experiment using loud speakers. As far as the results of this study are concerned, it is expected that the proposed method will work as a successful means to determine the priority among the multi-inputs in the source identification problems.
