Performance modeling and optimization of a smart sensor-based Hydroponic system using PSO and Genetic algorithms

Ensuring continuous and stable functioning of automated hydroponic systems is the key of sustainable agriculture, especially in resource-scarce regions or urbanization. This study explores the growing need for reliable smart farming technology by improving the reliability and availability of a hydroponic system which consists of sensors, micro controllers, and water pumps. A stochastic modeling approach using a Continuous-Time Markov Chain (CTMC) framework is adopted to simulate different system states namely operational, degraded, failed. The Chapman-Kolmogorov differential equations are derived from various state transitions and solved using Laplace transformation along with supplementary variable technique. These equations provide time-dependent probabilities, which are used to estimate the reliability and availability of the system. Nature-inspired optimization algorithms namely Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), optimize system performance and minimize downtime. The PSO yielded the highest availability of 0.9737 with the variation in iterations. Concurrently, GA achieved the highest availability of 0.9753 with population size 85. The findings show that using stochastic modeling alongside optimization techniques like PSO and GA helps in better parameter tunability for optimal maintenance planning and to understand how the system will behave over time.