PDDLMemory

PDDLMemory is an attempt to integrate a short-term memory system similar to those found in natural cognitive systems (i.e. human short-term memory) into classical AI planning. Classical AI planning deals with the problem of finding a plan (i.e. a sequence of valid actions or planning steps) within the space of possible world states (state space search) or that of valid plans (plan state search). The planning task consists of a logical description of the initial situation, a set of possible actions and a goal state that is to be achieved. Within a classical setting a planner searches for a solution to the problem (i.e. a valid plan) using all the information that is provided, i.e. everything which is known about the world. [more information]

https://github.com/pddlmemory

http://pddlmemory.github.io/pddlmemory/