Prediction Model And Application Of Stored Grain Temperature Based On BP Neural Network

As we all know,our country has a large population,but less arable land and limited food availability.Therefore,it is very important to ensure the safe storage of food.In the long-term storage of grain,temperature is one of the important factors to determine the quality of stored grain.When the temperature of grain exceeds the critical value of safe storage,it will cause a series of problems such as condensation,mildew,and insect damage.In order to ensure the quality of stored grain,it is urgent to grasp the temperature variation trend in grain pile accurately and make timely countermeasures.Therefore,it is of great practical significance to predict the temperature of grain storage comprehensively and accurately.Most of the existing predictions of stored grain temperature are based on the average temperature,but the single change of average temperature cannot represent the temperature change of the whole grain pile.In view of the above problems,considering the complex factors of the stored grain temperature,this thesis establishes a temperature prediction model based on the BP neural network to predict the maximum temperature of the stored grain.This thesis focuses on the following work:(1)According to the structure of the stored grain ecosystem,the factors affecting the quality of the stored grain are analyzed,including the temperature,humidity and moisture of the grain pile.On the basis of the research on the factors affecting the quality of stored grains,the variation rules of grain temperature,humidity and moisture are summarized.Combined with the study on the temperature change of the stored grain,the basic method of heat transfer and the basic law of heat conduction in the grain pile are analyzed.(2)The basic principles of BP neural network and Particle Swarm Optimization are described.Aiming at the problem that BP neural network optimized by Particle Swarm Optimization is easy to fall into the local optimal value,the BP neural network of Adaptive Mutation Particle Swarm Optimization is adopted in this thesis,and then this algorithm is applied to predict the temperature of stored grains.The BP neural network model optimized by Adaptive Mutation Particle Swarm Optimization algorithm is established to predict the maximum temperature of stored grain.The actual temperature data of the typical high and large warehouses are selected as examples,and the predicted results are compared with the BP neural network model and the BP neural network model optimized by Particle Swarm Optimization.The simulation results show that the BP neural network model optimized by Adaptive Mutation Particle Swarm Optimization algorithm has higher accuracy in predicting the temperature of stored grain.(3)Aiming at the temperature monitoring problem in the grain storage safety monitoring system,this thesis designs an Android-based stored grain temperature prediction module,which realized the temperature prediction of different grain layers in each granary in the system,and applies the temperature prediction module to the food storage safety monitoring system to provide information support for the mechanical ventilation.