Towards `human-consistent' multistage decision making and control models using fuzzy sets and fuzzy logic

Abstract Fuzzy multistage decision making (control) models are presented. Two ‘generations’ of them are distinguished. The first consists of conventional models derived from Bellman and Zadeh's general framework (fuzzy constraints, goals, decision, etc.). The second consists of newer approaches which additionally allow for a more general aggregation of control stage scores (outcomes), e.g., “… at most of the control stages…” instead of “… at all …” conventionally used. Both generations have roots in papers co-authored by Bellman. They are further steps towards human-consistent (soft) models whose implementation should be much easier.