Modeling the Dynamic Response of Plant Growth to Root Zone Temperature in Hydroponic Chili Pepper Plant Using Neural Networks

Abstract

One of the essential factors in the root zone environment that affects plant growth is temperature. Determining the optimal root zone temperature condition in a hydroponic system during cultivation could lead to an improvement in plant growth. An optimal control strategy can be determined by identifying the eco-physiological process using a dynamic model. However, it is difficult to develop a dynamic model of the responses of plant growth to root zone temperature because the eco-physiological processes of plants are quite complicated. We propose an intelligent approach that can deal with this complex system. Non-linear autoregressive with exogenous input (NARX) neural networks were used to develop a dynamic model of the responses of plant growth to root zone temperature. The responses of chili pepper plant growth as affected by root zone temperature were measured during 60 days of cultivation inside a growth chamber using a non-destructive and continuous system based on a load cell. Five datasets of dynamic responses of plant growth were obtained for system identification. The results suggest that the application of a neural network is useful for modeling the dynamic response of plant growth to root zone temperature in hydroponic cultivation, with promising performance.

Plant phenotyping relevance

植物成長を連続・非破壊に測定するロードセル系と、成長応答を推定するNARXニューラルネットワーク動的モデルが研究の中心であり、植物成長という形質の取得・モデル化手法に該当する。

Code and data availability

The paper cites the authors' own public Matlab program script for the NARX modeling of plant growth response to root zone temperature, hosted on GitHub (reference 47), which qualifies as a paper-specific public analysis code asset. No public phenotype dataset deposit is stated; the five measurement datasets are not明确ly