A feedback based CRI approach to fuzzy reasoning

Fuzzy reasoning methods are extensively used in intelligent systems and fuzzy control. In our previous work, an explicit feedback mechanism is embedded into optimal fuzzy reasoning methods, and the resulting fuzzy reasoning methods are more robust than the optimal fuzzy reasoning methods. An interesting question is that, without the introduction of optimization goals, can the robustness of fuzzy reasoning methods be improved by embedding feedback mechanisms? This paper is intended to answer the question. By embedding feedback mechanisms into CRI approach, a new feedback based CRI (FBCRI) approach is proposed and three implementation methods with different strategies are obtained. Simulation results show that the feedback mechanisms are really useful for the improvement of the robustness of CRI methods. At last, to test the applicability of the proposed approach, it is applied to the solution of a real-time path planning problem for UAVs. The effectiveness and efficiency of an FBCRI based real-time path planning algorithm are verified by representative test examples, which show that embedding feedback information into the fuzzy reasoning process actually improve the quality of flight paths.