Stored Grain Pest Activities Acoustic Signal Characteristics Study

Our country is a big agricultural country, the storage of grain is not only related to the need of military and people, but also related to the national security and social stability, so the storage of grain has great importance.The stored grain is often damaged by pests. The data from the Food and Agriculture Organization of the United Nations show that the loss of the grain in the whole world accounts for5%of the total each year, and if the manual labor, material and technology can't meet the need of grain storage, the loss may account for20-30%. As a big country in grain production, storage and consumption of the world, China's grain loss during storage is up to8%~10%, which makes the storage of grain more and more urgent. The enhancement of grain storage technology is imminent and early detecting of species and quantity of the stored grain insects has great significance.In various methods of detecting stored grain insects, the method based on activity sound of stored grain insects is favored, because it is light, environmentally sustainable, simple, nondestructive, fast and flexible.The research designs the apparatus of activity sound sampling of insects in stored grain. The methods of power spectrum analysis and wavelet packet decomposition are respectively used to research the characteristics of insect activity sound in stored grain. The sound characteristics at different temperatures are deeply made comparisons and analysis. The experimental results provide data for establishment of sound characteristics database of stored grain insects and the achievement of insect automatic detection.The research mainly includes the following aspects:(1) The research background and significance of activity sound of stored grain insects and current development of detection methods of stored grain insects are summarized. In particular, the foreign and native research findings on detection methods of activity sound of stored grain insects are mainly summarized.(2) The morphology and behavior characteristics of maize weevil and rice weevil are introduced, moreover, the system of activity sound sampling of insects in stored grain is proposed. Its hardware devices, settings of computer sound recording software and the particular process of activity sound sampling are all presented. (3) After respectively sampling the activity sound of rice weevil adults, maize weevil adults and rice weevil larvae in wheat, corn and coix seed, the power spectrum characteristics of activity sound of these insects were analyzed and differences at different temperatures were studied. The conclusion shows that at27℃, the frequencies of activity sound power spectrum peaks of rice weevil adults in wheat are172.3Hz,409.1Hz,1077.OHz, the frequencies of activity sound power spectrum peaks of maize weevil adults in wheat are193.8Hz,430.7Hz,990.5Hz, the frequencies of activity sound power spectrum peaks of rice weevil adults in corn are236.9Hz and1012.OHz, the frequencies of activity sound power spectrum peaks of rice weevil larvae in coix seed are344.5Hz and990.5Hz. On the other hand, the conclusion shows that both insect species and grain species affect the activity sound characteristics of stored grain insects, and the frequency ranges of activity sound power spectrum peak at different temperatures are different.(4) The wavelet packet decomposition characteristics of activity sound of stored grain insects are studied. The characteristics include energy of node,energy ratio of node and energy entropy of signal.The energy of node,energy ratio of node of activity sound from maize weevil adults in wheat and corn at the same temperature are respectively compared. Getting the bands which have characteristic differences in this temperature. The study not only gets statistical regularities of average of energy entropy of activity sound from maize weevil adults in wheat and corn at different temperatures, but also gets statistical regularities of average of energy,average of energy ratio of WP decomposition nodes and average of energy entropy of activity sound from rice weevil larvae in coix seed at different temperatures.