Design Of Storage Grain Temperature Prediction And Safety Warning System Based On GWO-SVR

In the process of grain storage,the temperature of grain storage exceeds the critical value of safe storage,and the phenomenon of dew,mildew and insect damage is one of the important causes of grain storage loss.To ensure the safety of grain storage,it is especially important to accurately grasp the trend of grain pile temperature changes and take corresponding disposal measures,and the study of grain situation early warning methods has certain practical significance for the safe management of grain storage.Therefore,the BIM based grain storage safety early warning system is constructed to realize the three-dimensional visualization management of grain silos,predict the temperature of grain piles with the help of machine learning technology,and combine with mechanical ventilation technology to realize intelligent ventilation of grain silos,so as to provide a basis for the decision making of grain silo managers,thereby reducing grain losses in the process of grain storage and improving the safety management level of grain silo buildings.In order to establish an effective early warning system for grain storage security,the following work is done:(1)Firstly,the selection of grain temperature prediction models is based on the concept of machine learning,secondly,a grain silo grain data is selected,and the missing and abnormal cases in the data are processed manually,and three models of random forest,support vector regression and BP neural network are constructed for grain storage temperature prediction based on the processed data set.The model with the best prediction effect was determined by combining the comparison graph of prediction results and evaluation index.(2)The gray wolf algorithm was used to perform parameter search for the support vector regression model with the best prediction results,and a GWO-SVR hybrid prediction model was established.The prediction results of GWO-SVR and SVR prediction models based on the same data set are compared,and the results show that the support vector regression model optimized with the gray wolf algorithm has higher accuracy and improves the prediction accuracy of the SVR model,and it is used for the establishment of the subsequent early warning system.(3)Firstly,the demand analysis of grain storage safety early warning system is carried out,followed by reviewing safety grain storage standards and relevant literature to determine the critical value of safety grain storage temperature,and combined with the prediction results of GWO-SVR model,reasonable ventilation measures are suggested for dangerous grain conditions.Finally,combining the means of grain situation early warning with BIM technology,a city grain reserve warehouse is taken as an example to build its BIM model,and the detailed design of grain situation detection,safety warning and intelligent ventilation function modules is carried out,and some functions are demonstrated.