A Prototype Simulation Based Optimization Approach to Model and Design an Advanced Life Support System 2004-01-2576

In this paper a SIMulation based OPTimization (SIMOPT) approach is used to study the dynamics of Advanced Life Support System (ALSS). The SIMOPT architecture uses a Deterministic Optimization (DO) algorithm to optimize the overall ALS behavior by minimizing the re-supplies which are difficult to procure or transport, in conjunction with a simulation model which introduces uncertainty, i.e. randomness, to the system. DO algorithm is a detailed deterministic optimization model of ALSS, which is used to determine the values of strategic decisions, such as the crop growth area. An aggregate time-dependent mass balance model of ALSS and an aggregate steady state mass balance model of ALSS are developed as the simulation and optimization modules in SIMOPT, respectively. The ranges of acceptable values of strategic decisions, e. g. safety buffers for oxygen, edible food, water and carbon dioxide, in a given ALSS scenario are determined using SIMOPT which utilizes time series data mining methods.