Forecasting of Energy Requirements for Planetary Exploration Habitats Using a Modulated Neural Activation Method
Abstract
Human travel to other planetary bodies will have duration of years, with little to no possibility of resupply. Consequently, the monitoring and forecasting of resource consumption is a mission critical capability. The actions of a single crew member can heavily influence small environments and throw forecasting models off to the point of total failure. It is hypothesized that the inclusion of variables accounting for daily astronaut activities and psychological state will allow for higher accuracy in forecasting. Utilizing consumption data from the Hawaii Space Exploration Analog Simulation in the form of electricity, and psychological data from the Positive and Negative Affect Schedule, a generalized artificial neural-modulation method is introduced which allows the incorporation of emotional response into machine learning applied to forecasting. It is found that the changes in the activation functions correlate to observed crew behavior, and show significant improvement of forecasting results.
Description
Keywords
Artificial Intelligence
Citation
Engler, S. (2017). Forecasting of Energy Requirements for Planetary Exploration Habitats Using a Modulated Neural Activation Method (Master's thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca. doi:10.11575/PRISM/26208