摘要：Cow's rumination is an important process of food digestion for dairy cows which is a valuable indicator during the cow management. By measuring the cow's rumination time, it is capable to predict the cow's estrus, and learn the cow's health situation. This paper uses MSP430F149 processor with a sound sensor to achieve the design of real-time monitoring system for cow's rumination. Since the system require high accuracy of the audio signal of the cow's rumination, the ADC acquisition module and filtering and amplification process was made a special design. For the change of the spectral characteristics of the voice when cows ruminating, this paper designed the endpoint detection algorithm and sound sequence windowed function, and FFT transform is performed on the data in the Hamming window, and then do the frequency domain analysis to the audio of cow's rumination. Comparing the sound spectrum collected from high precision recording instrument and this system, this system has done a good job in frequency and time domain. Applying the high precision recording instrument to the simulation experiment for cow's rumination which conduct 50 sets of different time ranging, using this system to do the real-time monitoring, the mean error was 4.38, the R 2 value is 0.877, the Pearson correlation was 0.936. This system performed accurate in the acquisition of the sound signal, and more accurately in identifying cows ruminating, and it can also provide effective means for the intelligent management of the pasture on cow's health and breeding.
关键词：cow's rumination; real-time monitoring; endpoint detection; window ; function; time and frequency domain
