摘要：It is difficult to accurately predict the photosynthetic rate of crops in the greenhouse, which affects the implementation of greenhouse light environment regulation considering photosynthesis demand. Therefore, this paper takes tomatoas the test crop, and through the combination experiment onphotosynthetic rate with the condition of nested temperature,humidity, photon flux density and CO2 concentration, thesample data is obtained. Then, the training samples are selectedfrom the sample data to train the SOPSO-LSSVM algorithm,and a soft sensing model for predicting the photosynthetic rate isestablished, which can realize the accurate prediction of tomatophotosynthetic rate. Finally, the simulation results indicatewhen the model is used to predict the photosynthetic rate, ithas the smaller maximum relative error and root mean squareerror, which are 0.0264 and 0.2926 respectively. The averagerelative prediction error is only 0.0078 and the coefficient ofdetermination is 0.9953. More shows that the established modelcan provide a reliable objective function for the regulation ofgreenhouse light environment considering the photosynthesisdemand, and then guide the regulation of light environment.
