Research On Grain Storage Measurement And Control System Based On Grain Pest Prediction

Food security is the foundation of national security.In recent years,China has put forward a clear plan for the informatization and intelligent transformation of grain storage,and "scientific grain preservation" has become an important industry in the grain industry.However,there are still nearly 5% grain storage losses in China.The main reasons for these losses are the respiration of grain itself,the loss of water and the consumption of stored grain pests,among which the consumption of stored grain pests accounts for the majority.Therefore,scientific and effective control of stored grain pests is of great significance to China’s food security.At present,most of the researches on stored grain pests are based on image recognition of pests.In recent years,there have been studies on the prediction of grain pests,but many related studies only consider the temporal characteristics of pests,ignoring the influence of temperature,humidity and other granary environmental factors.Similarly,the current grain storage measurement and control system on the market lacks a functional module integrating detection,prediction and early warning of grain storage pests,which cannot fundamentally solve the outbreak of grain storage pests.In this paper,the pest prediction model of three-layer LSTM stack network was used for reference and optimized based on the time characteristics of storage grain pest collection and the time series data of temperature and humidity.Considering the spatial characteristics of temperature and humidity sensors and pest traps in grain stacks,the CNN network was introduced to extract the spatial characteristics,and the three-layer LSTM stack network was further optimized.In addition,considering the connection between the forward and backward direction of time series data,the bidirectional LSTM structure is introduced,and the network model of CNN + double-b-LSTM is optimized.Finally,combined with the data collected in the field of grain depot,the experiment verifies the effectiveness of the three optimization algorithms.In addition,the defects in the pest detection,prediction and warning functions of the grain storage control system on the market are also discussed.This paper designs and implements an intelligent monitoring and control system for stored grain pests.The whole system adopts the design mode of front and rear end separation,which can monitor and predict the stored grain pests,and generate pest reports for early warning.While participating in the project research and development of the overall grain storage measurement and control system,this paper combined with the key chain technology of project process management to optimize the project schedule.While ensuring the quality of products,the project duration was greatly shortened and the benefits were brought to the cooperative enterprises.