Research on Temperature Prediction Model in Greenhouse Based on Improved SVR

Abstract

The greenhouse is a multivariate, nonlinear, time-varying and time-lag system. The establishment of an accurate greenhouse temperature prediction model can effectively ensure the effectiveness of intelligent temperature control. In order to solve the problem of difficulty in modeling the temperature mechanism of greenhouses, this paper proposes a method based on improved Support Vector Regression to establish greenhouse temperature prediction, The PSO algorithm is used to optimize the hyper-parameters of the SVR. Using real greenhouse data to verify the effectiveness of the proposed model. Experimental results show that compared with the traditional BP neural network prediction model, the temperature prediction model in greenhouse based on improved SVR has significantly improved prediction accuracy and better prediction capabilities. Therefore, the greenhouse temperature prediction model based on improved SVR proposed in this paper can well characterize the actual greenhouse system.

Prediction of Internal Temperature in Greenhouses Using the Supervised Learning Techniques: Linear and Support Vector Regressions

“…Ge et al [35], compared various models; specifically, the SVR model had lower precision. Fan et al [33], had similar results using SVR BRF but only evaluated the error using the MSE metric; hence, more information is needed to confirm the model's accuracy.…”

“…In refs. [32][33][34] use SVM to make predictions about certain variables in greenhouses. Other studies used algorithms such as Xgboost [35] or LSTM [36] to analyze meteorological factors affecting crop evapotranspiration.…”

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

“…However, BP neural networks often stagnate in the flat area of the error gradient surface, and their convergence is slow or potentially unable to converge. Fan et al [16] used the PSO algorithm to optimize support vector regression (SVR) super parameters, the results of which indicate that the improved SVR greenhouse temperature prediction model outperformed the classic BP neural network. Yuan et al [17] combined the PSO algorithm with the LS-SVM algorithm, which has higher accuracy and performance for predicting photosynthetic rates than the traditional BP or SVM algorithms.…”

A simple wall in front of the greenhouses

“…The problem is visualised with technologies for thermal, luminescent and humidity control (Kavga et al, 2020;Kokieva et al, 2020) as thermo-physical parameters (Du et al, 2021;Fan et al, 2021) and infrastructure designs (Yang, 2021) including Internet of Things technology (Han et al, 2018). To these technologies can be added studies of the effects of wind shear by high but low-cost constructions (Jani et al, 2021;Li et al, 2018).…”