In the process of damage detection in communication electronic circuits,the accuracy of damage detection is low due to some microdamage features that are difficult to detect.Therefore,a machine learning algorithm based microdamage detection method for communication electronic circuits is proposed to achieve accurate and efficient detection of microdamage in electronic circuits.Introducing wavelet packet feature analysis,extracting the kinetic energy features of wavelet signals,calculating damage sensitivity,and constructing a communication electronic circuit microdamage detection model based on one-dimensional convolutional neural network.The extracted kinetic energy features are used as model in-puts,and the feature samples are distinguished and labeled,and the training data set is put back for sampling to achieve model parameter updates.Finally,the electronic circuit microdamage detection is achieved by inputting the tested circuit samples into the optimal training model.The experiment shows that the proposed algorithm effectively detects microdamage in communication electronic circuits,has high applicability and feasibility for micro damage detection.and has strong anti noise interference ability.
维普
万方数据