On-site verification of gravel mass estimation method with hydrophone
2018
Takaki, K. (Institute for Rural Engineering, National Agriculture and Food Research Organization, Tsukuba-shi, Ibaraki (Japan)) | Mukai, A. | Tanaka, Y. | Tsunesumi, N.
We verify the gravel mass estimation method with a hydrophone in actual agricultural channels. In this estimation method, several models are created to classify the gravel mass into two classes, a large class and a small class. However, it is difficult to utilize the data acquired for identification on the site where various noise generation is conceivable. Therefore, we assumed that the number of pulses obtained by collision of gravel is represented by a unimodal probability distribution of the histogram with the sound pressure as the class and the pulse number as the frequency, and the data necessary for modeling was extracted. Moreover, in consideration of applicability to the site, the models were applied by removing the data assumed to contain background noise and multiple collision noise, regardless of the mass of the target gravel. As the result, good mass classification by the proposed models was confirmed, even in the site under the influence of noise all the time.
显示更多 [+] 显示较少 [-]