Analysis of Dynamic Mode Decomposition of Temperature and Relative Humidity Time Series in a Simulated Greenhouse

Accurate control of greenhouse microclimates is crucial for optimizing crop production. Simplified models, such as the perfectly stirred tank approach, are often used as a valid assumption for understanding the complex system of greenhouses. This study applies Dynamic Mode Decomposition (DMD) to analyze time series temperature and relative humidity data in a simulated greenhouse environment. Data were generated over ۳۰ days, and the first day was eliminated from the data set, assuming homogeneous mixing within the greenhouse. The algorithm was implemented in C++ using the Eigen ۴.۳.۰ and VTK libraries and Visual Studio as an integrated development environment (IDE). The DMD reconstructed the data with a relative error of less than ۱۰^-۱۳, which indicates the ability of this method in time series analysis.

کلیدواژه ها:

Dynamic mode decomposition ، Time-series analysis ، Greenhouse modeling ، Data-driven ، Reduced Order Models