Nuclear Electric Vehicle Optimization Toolset (NEVOT)

The Nuclear Electric Vehicle Optimization Toolset (NEVOT) optimizes the design of all major nuclear electric propulsion (NEP) vehicle subsystems for a defined mission within constraints and optimization parameters chosen by a user. The tool uses a genetic algorithm (GA) search technique to combine subsystem designs and evaluate the fitness of the integrated design to fulfill a mission. The fitness of an individual is used within the GA to determine its probability of survival through successive generations in which the designs with low fitness are eliminated and replaced with combinations or mutations of designs with higher fitness. The program can find optimal solutions for different sets of fitness metrics without modification and can create and evaluate vehicle designs that might never be considered through traditional design techniques. It is anticipated that the flexible optimization methodology will expand present knowledge of the design trade-offs inherent in designing nuclear powered space vehicles and lead to improved NEP designs. I. Introduction HE integration issues associated with using reactor power for actual space science missions have received little attention to date and the design trade-offs that must occur on a reactor-powered space vehicle are not fully known. The Nuclear Electric Vehicle Optimization Toolset (NEVOT) is designed to uncover the unknown trades so they can be studied and understood. NEVOT is a nuclear electric propulsion (NEP) vehicle optimization algorithm based on a genetic algorithm (GA) search technique. The tool searches for the optimal vehicle subsystem combinations to perform user defined missions. Traditional methods of multi-disciplinary system design are sequential in nature. A subsystem is selected and designed, and inputs from the first subsystem design are passed to the "next" subsystem, which is sized or designed to be compatible with the first subsystem. The process is repeated until all of the subsystems are designed. The entire system is then compared to known constraints and compromises are made between subsystems to meet those constraints. Once a base design is complete, the system is perturbed in an effort to optimize it for a desired result, such as minimizing mass or cost. Automated sequential multi-disciplinary design efforts typically mimic human designers and are sequential in nature. The resulting configurations are a strong function of the information design logic built into them. NEVOT attempts to remove potentially limiting foreknowledge from the design process in order create subsystem arrangements that might never result from traditional design methods, whether automated or not. By removing the trends that are inherently designed into complex multi-disciplinary systems we can hopefully uncover new trends that will lead to improved designs. The tool allows for the combination of tens of thousands of subsystem designs without regard to their compatibility or combined ability to meet the mission requirements. The ability of each combination of subsystems to perform a mission is evaluated after they are created and assigned a