Measurement and Control of Water Status of a Plant (Part 2)

Identification of Autoregressive Models

Abstract

Control of water status of a plant is important for control of crops. In this report, cross correlation of water potential with vapor transpiration and diameter variation were discussed. Autoregressive models were assumed for the plant's response models, and parameters were identified from the plant's responses to randomly varying light intensity. In the simulation, step responses were calculated, and the results showed that the transpiration response corresponded to the model. Autoregressive spectral analysis was also done, and the result showed it was more precise than spectral analysis by FFT method. Identification of autoregressive models showed good results for evaluation of control models and estimation of spectral characteristics.