The Abnormal Sound Detection Mechanism Research Of Loudspeaker And Its Implementation Method

The abnormal sound detection technology of loudspeakers is vital for their quality,so the first step is pure tone detection after the production of loudspeakers.This traditional method mainly relying on the manual detection is influenced by various subjective factors can't meet the needs of online production.Many computer-based detection technologies for loudspeakers,including various hardware devices and supporting development of software technology,still have shortcomings,such as:low detection efficiency,low accuracy,high cost and so on,resulting in online production of electro-acoustic devices is difficult to achieve the technical requirements of high yields,high quality and efficient detection.In the view of all the defects,an abnormal sound detection method based on signal processing technology for online loudspeakers is proposed in this paper to further improve the accuracy and efficiency of the algorithm for detection system,which achieves the technical requirements of high performance,high precision and easy operation for online automatic detection suitablyThe study of algorithm in this paper mainly includes three aspects.firstly,the improvement of the psychoacoustic model resolution algorithm based on different window functions,which is used to improve the efficiency and accuracy of abnormal sound detection system based on psychoacoustic model;Secondly,the combination of digital image technique and the Hilbert spectrum analysis for the feature extraction algorithms of the abnormal sound;Finally,the research of algorithm based on empirical mode decomposition algorithm for the abnormal sound detection.The findings are as follows:(1)The study of the psychoacoustic model resolution algorithm based on different window functions improves the resolution of the psychoacoustics model effectively,which is applied to the abnormal sound detection system based on the psychoacoustic model to further improve the accuracy and efficiency.(2)The comparison the advantages and disadvantages of the different time-frequency analysis method is studied for the noise detection.By the end,the decomposition and detection for the noise signals based on the instantaneous frequency and instantaneous amplitude are achieved successfully by using the Hilbert-Huang algorithm(3)The digital image binarization processing technology is applied to the Hilbert spectrum analysis,which extracted the features of the common types of abnormal sound remarkably both in the time-frequency domain and the three dimensional spectrum.It can be used as the basis of online fault detection for the loudspeaker.(4)An abnormal sound detection and extraction algorithm based on empirical mode decomposition algorithm is developed,which is verified by simulations for ensuring its effectiveness.In conclusion,the above improvement algorithms for abnormal sound detection system are verified through a lot of the simulations,the results demonstrate the effectiveness and feasibility of the proposed algorithms.84charts,6 tables,69references.