Conformant Planning based on Approximation (CpA)
Description The two planner systems (CpA(C) and CpA(H)) are the successors of the planner CpA [STGM05]. The basic idea of these systems lies in the use of approximation reasoning, in particular the so called 0-approximation, in the search for a plan. This method reduces the size of the state-space from 2^{2^n} to 3^n. CpA(C) and CpA(H) differ from CpA in that they are complete for domains without static causal laws. Both systems uses a combination of three heuristics: cardinality, length of relaxed plan, and number of satisfied subgoals as their heuristic function. CpA(C) prefers cardinality while CpA(H) uses length of relaxed plan as its primary heuristic function.
Binary code The binary code can be download here. CpA(C) and CpA(H)
Preprocessor Source code (need to run with Sicstus or SWI-prolog)
Translator PDDL to PL
Test domains (Each tar file contains the inputs for a planner. The name of the tar file indicates the planning system. PDDL is used for both CpA(C) and CpA(H).