ANALYTICAL MODELING OF MICROCLIMATIC PARAMETERS OF THE PLANT ENVIRONMENT

Abstract

The paper presents an intelligent system for automated monitoring and forecasting of microclimatic parameters for caring for houseplants. The developed system provides continuous monitoring of soil moisture, air temperature, and atmospheric pressure with the possibility of short-term forecasting based on machine learning methods. The basis of the mathematical model is a modified Ridge regression with regularization, which increases the resistance to multicollinearity of input parameters and ensures the accuracy of forecasting with significant data fluctuations. To describe daily and seasonal changes, feature expansion was used, taking into account cyclic components that reflect the natural periodicity of evaporation and cooling processes. The experimental database contains 3764 hourly records obtained during five months of observations. The range of soil moisture was from 30 to 80%, the temperature varied from 20 to 28 °C, and the atmospheric pressure was within 1003–1023 hPa. According to the modeling results, the average error in the humidity forecast does not exceed 2.7%, which confirms the high accuracy of the algorithm. The conducted correlation analysis revealed a strong inverse relationship between temperature and humidity (r ≈ -0.65) and a high direct correlation between the water stress index and the evapotranspiration rate (r ≈ 0.85), which reflects the physiological consistency of transpiration processes. The results obtained allow to automatically determine the optimal moment of irrigation, reducing water consumption by 25-40% compared to traditional threshold systems. The proposed technology combines sensor measurements, analytical algorithms and adaptive control, which ensures the transition from reactive to proactive microclimate management. The practical significance of the study lies in the formation of a scientifically based concept of a “smart flowerpot”, aimed at increasing irrigation efficiency, stability of growth conditions and sustainable development of home plant growing