Design And Implementation Of Intelligent Grain Condition Monitoring And Early Warning System

Now,the safety management of grain storage is still partially or completely dependent on manpower,So grain storage enterprises urgently need a set of intelligent measurement and control system.Thus,thesis design and implementation an intelligent grain condition monitoring and early warning system,which can find and solve the problems in grain storage.Firstly,thesis analyzes the requirements of the system,and then design and implementation the whole and each module.Finally,thesis test the system.Technically,a set of Java Web system is developed based on B/S mode.Vue progressive framework is used in the front end,Spring boot framework is used in the back end,SQL Server database is used.At present,there are still many problems in the field of grain situation monitoring and early warning: first,the grain security judgment only considers the internal storage of grain and ignores the impact of the external environment;second,PSO algorithm optimize the RBF neural network algorithm commonly used in grain temperature prediction,but the local optimization ability of PSO algorithm is still poor;finally,The number of keywords extracted in semantic query of grain situation cannot be adjusted dynamically according to the query statement.(1)According to the judgment of grain safety,thesis considering the influence of the external environment,Calculate the relative temperature and humidity by weighting the three temperature and three humidity respectively,and the membership degree of relative temperature,relative humidity,grain moisture are used as the input of neural network.(2)According to the problem that PSO algorithm is easy to fall into local optimum,thesis combines genetic algorithm with PSO algorithm,and calls genetic algorithm to change particle position when PSO algorithm falls into local optimum.At the same time,through comparative test,it is determined that the prediction effect is better through the data of the previous days.(3)According to the query characteristics of the internal database in the grain situation semantic query system,thesis proposes a method to dynamically change the number of key words extracted according to sentence pattern and part of speech.After multiple use tests and comparative tests,The results show that this method improves the performance of the original algorithm and achieves the expected goal.