In order to meet the engineering requirements of non-contact measurement of impact load and realize the intelligent identification of its characteristic parameters, this paper attempts to establish a method to analyze and learn the audio characteristics generated by accidental loads, such as impact and explosion, and to identify the impact load characteristics in reverse. The method, which is based on energy conservation, momentum conservation and strokes law, used p-norm in mathematics to extract the basic characteristics of impact load, and the frequency analysis to find the inevitable connection between the audio characteristics impulse-induced and the impact load characteristics. At last, “Acoustic Discrimination for Impact Load Identification (ADII)” by adopting the functions of artificial neural networks about learning, categorization and generalization was implemented.
Paper
Full text
An Acoustic Discrimination Method for Impact Load Based on Artificial Neural Network
Semantic Scholar · Engineering · 2020
Abstract
In order to meet the engineering requirements of non-contact measurement of impact load and realize the intelligent identification of its characteristic parameters, this paper attempts to establish a method to analyze and learn the audio characteristics generated by accidental loads, such as impact and explosion, and to identify the impact load characteristics in reverse. The method, which is based on energy conservation, momentum conservation and strokes law, used p-norm in mathematics to extract the basic characteristics of impact load, and the frequency analysis to find the inevitable connection between the audio characteristics impulse-induced and the impact load characteristics. At last, “Acoustic Discrimination for Impact Load Identification (ADII)” by adopting the functions of artificial neural networks about learning, categorization and generalization was implemented.