Estimation method for warehousing grain temperature field based on transfer learning

Abstract

The invention discloses an estimation method for a warehousing grain temperature field based on transfer learning. The method comprises the steps that a mixed-effect model framework combining a thermodynamic model with a transfer learning model is established, a temperature value ym (s,t) of the three-dimensional grain temperature field at a space-time point (s,t) is expressed as a sum of an overall temperature variation item of the temperature field, a local temperature variation item of the temperature field and a noise item caused by random or uncontrollable factors, overall change of the grain temperature field is described and estimated through the thermodynamic model, the transfer learning model of a space-time temperature field is established and local change of the grain temperature field is described and estimated by combining a multi-task learning method with a self-correlation time sequence method, the warehousing three-dimensional grain temperature field is accurately estimated, and therefore the warehousing grain temperature is obtained. According to the method, the problems can be solved that sensor data is insufficient, observation data is partially lost, and information of the grain temperature field cannot be accurately obtained; the comprehensive and accurate information of the grain temperature field is obtained.

Description

technical field [0001] The invention provides a method for estimating a temperature field, in particular to a method for obtaining stored grain temperature by estimating a grain temperature field based on transfer learning, and belongs to the field of industrial engineering. Background technique [0002] As the world's population increases, so does the demand for food worldwide. In order to meet the supply of grain, the Chinese government has built a large number of grain depots across the country to store grain. During the storage process, grain will deteriorate due to various reasons, which will eventually lead to the loss of grain. The loss of grain in storage has become a widespread concern around the world. According to the report of the Food and Agriculture Organization of the United Nations, about 8% of the food in the world is damaged every year due to the existence of pests. It is very important to monitor the grain quality in the storage process, and the most co...

AI Technical Summary

Problems solved by technology

In order to solve the problem of insufficient sensor data in the current granary, the present invention adopts transfer learning technology and simultaneously considers multiple granary sensor data with similar attributes to realize accurate estimation of the three-dimensional grain temperature field in the storage process

Images