Prediction Model of Carbon Dioxide Concentration in Pig House Based on Deep Learning

The air environment (e.g., high concentration of carbon dioxide) in a pig house will affect the health conditions and growth performance pigs, quality pork as well. In order to reduce cumulative dioxide house, prediction model was established by deep learning method predict changes house. This also be used for real-time monitoring adjustment experiment designed collect environmental parameters temperature, humidity, wind speed, concentration) data several months. ensemble empirical mode decomposition–gated recurrent unit (EEMD–GRU) results show that compared with other models, accuracy EEMD–GRU is highest, root mean square error (RMSE), absolute (MAE), percentage (MAPE), r-squared (R2) autumn winter are 123.2 ppm, 88.3 3.2%, 0.99, respectively. RMSE, MAE, MAPE, R2 129.1 93.2 5.9%, 0.76 spring summer. proposed this paper can effectively provide effective help precise control environment.