| dc.description.abstract | Water scarcity is a global challenge with only one-third of the world’s population having consistent access to clean drinking water. Atmospheric water harvesting is a promising approach owing to the significant amount of water, i.e., 13000 trillion liters, present in the atmosphere. While significant recent research has focused on developing innovative sorbent materials, components and system designs, there is limited understanding of how to optimize device performance through active control. Key operating parameter selection, specifically, desorption temperature and cycle length, has relied on experimental trial and error. In this thesis, model predictive control (MPC) was used to dynamically optimize power input and cycle time in atmospheric water harvesting devices, for the first time. Real time optimization using a custom defined cost function was achieved based on a simplified heat and mass transfer model. The model allowed for the cost function to be based on water output and therefore eliminated the need for setpoint definition a priori. Through a modular, customizable software and hardware stack, the device demonstrated reliability and maintainability while preserving user interaction. MPC was evaluated against five distinct sorbent isotherm types, using three distinct operating modes: maximizing water production, maximizing operational profit and increasing thermal efficiency. All modes outperformed a constant temperature setpoint by dynamically determining the appropriate end time of the cycle, which depending on the material varied up to 10,000 s. Furthermore, the controller was able to increase thermal efficiency up to 3 percentage points compared to the reference by dynamically tapering power input to match water production. Experimental validation was performed with a device built by the Device Research Laboratory. The results showed excellent agreement between measured water output and real-time prediction, which provides a viable strategy for future controller deployment. This work paves a way for more sophisticated device operation through real-time optimization of power input and cycle length and highlights a modular software and hardware design to realize high performance atmospheric water harvesting devices. | |
