Evolutionary Algorithms for Planning in an Autonomous Agent
Christoffer Dahl · NORA - Norwegian Open Research Archives · 2010
The planning of constraint-based operations using evolutionary algorithms is an optimization problem that involves finding a chain of operations appearing in a valid left-right or right-left order. This type of planning can be applied to autonomous systems where the need to plan actions for one or several agents is relevant. Current research focuses on developing autonomous management systems for the use on oil drilling facilities, aiming to replace human-based micromanaged decisions with artificial intelligent solutions. This thesis describes one approach for planning sequences of constraint-based operations formed on an evolutionary algorithm theory.