Methodology for Optimal Plant Growth Strategies in Life-Support Systems

Abstract

A controlled ecological life-support system (CELSS) combines the photosynthetic activity of plants with physico-chemical processes to regenerate waste material into food and oxygen. The sustainability of such a semiclosed system requires that a controller making strategy decisions has a variety of mass flow options from which to choose. A methodology is developed to determine plant growth strategies for CELSS candidate species. Strategy decisions include the desired biomass at the time of harvest, the proportion of inedible biomass hydrolyzed to sugar and used as a food source, and the amount of phosphorous in the plant nutrient stream. These studies suggest ways in which optimization can be used to increase understanding and establish design criteria. A Monte Carlo simulation technique is combined with a nonlinear simplex optimization algorithm to find a range of strategies producing acceptable diets while minimizing resource cost. Whereas random guessing of potential strategies produced only 53 solutions in 1,000,000 trials, our methodology found acceptable low cost diets in 40–60% of trials. This efficiency is important in providing the controller with a range of viable operating points.