An accurate disturbance source locating method based on machine learning for complex environments

A method for disturbance positioning in noisy environment is introduced. By the partitioning algorithm and a well-trained B-P neural network classifier, an average locating accuracy of 95% was demonstrated in the field test.