Traffic noise automatic identification method based on ACF and IACF

Abstract

The invention discloses a traffic noise automatic identification method based on ACF and IACF. The method includes the steps that (1) noise generated when a vehicle normally runs is collected; (2) acoustical signals collected in the step (1) are converted into digital signals, ACF and IACF function computational analysis is carried out, and therefore index parameters of all sampled points are obtained, wherein the index parameters include the index parameter phi (0), the index parameter pi1 and the index parameter pi1ACC; (3) feature extraction is carried out on the index parameters, wherein a characteristic value P of the index parameter phi (0), a characteristic value s of the index parameter phi (0), a characteristic value R of the index parameter pi1 and a characteristic value Ct of the index parameter pi1ACC are included; (4) classification identification is carried out on the acoustical signals. Compared with a traditional method that a sound pressure level and a frequency spectrum are used for noise measurement, the method solves the problems that the difference between a test and actual hearing experience of people is large, measuring result errors are large, and real-time performance and representativeness of obtained monitoring data are poor.

Description

technical field [0001] The invention belongs to the field of environmental noise monitoring, in particular to an automatic analysis and identification method for road noise, railway noise and noise caused by different train types. On the basis of the method of the invention, an automatic noise monitoring system can be established for automatic identification and automatic monitoring of environmental noise (especially traffic noise). Background technique [0002] Environmental noise monitoring is the basis and theoretical basis for urban noise control, and it provides first-line data for noise policy formulation and noise control. For a long time, sound pressure level (including equivalent sound pressure level, maximum sound pressure level and cumulative percentage sound level, etc.) and frequency spectrum have been the physical indicators corresponding to subjective evaluation in noise measurement. They are also the reference standard for noise policy development. GB3096-2...

AI Technical Summary

Problems solved by technology

In this way, it can solve the problems of using traditional sound pressure level and spectrum to measure noise and the actual listening experience of human beings, the error of measurement results is large, and the real-time and representativeness of the obtained monitoring data are poor; automatic monitoring replaces manual Monitoring, in addition to effectively solving the problem of human factors affecting the accuracy of measurement results in manual monitoring, it can also facilitate the actual monitoring work and save a lot of human resources

Images