Predicting Multidimensional Environmental Factor Trends in Greenhouse Microclimates Using a Hybrid Ensemble Approach

Trend prediction of greenhouse microclimate is crucial, as crops are vulnerable to potential losses resulting from dramatic changes in microclimate. Consequently, a precise predictive model required that can predict trends microclimates several weeks advance avoid financial losses. In the present study, we proposed hybrid ensemble approach based on an Informer optimized using improved empirical mode decomposition (IEMD). The dataset was decomposed IEMD, and then all datasets were predicted model. Afterward, predictions combined. five different environmental factor CO2 concentration, atmospheric pressure, light intensity, temperature, humidity predicted. performance IEMD-Informer compared with other modeling approaches. results demonstrate method has outstanding factors more accurately.