Embedding inference engine in fuzzy expert robotic system shell in a humanoid robot platform for selecting stochastic appropriate fuzzy implications for approximate reasoning

The purpose of this research is to select an appropriate fuzzy implication for approximate reasoning under each situation and to solve reasoning based on fuzzy production rules, which is usually referred to as approximate reasoning in the design of inference engines for fuzzy expert systems, Predicting the knowledge domain is removed from an expert system, the remaining structure is to extract an expert system shell, Here the applicability of an expert system shell is not necessarily restricted to one particular knowledge domain. An inference engine embedded in an appropriate expert system shell is made reusable for different domains of knowledge for different expert systems, with relevant human experts. In this research we identified meaningful criteria in terms of which distinct fuzzy implications could be evaluated and compared. In this research the explanatory interface facilitates communication between the user and the expert system. Approaches to evaluation are done for the relevant production rules by modus ponens. In this research it has been observed that experienced results are able to perform and recognized that the areas of fuzzy systems and neural networks are strongly interconnected, Here in this research neural networks have been proven that fuzzification is very useful in this humanoid robotic research using fuzzy set for constructing membership functions of relevant fuzzy sets and other context-dependent entities from sample data. We had explored here that the motivation for approximating fuzzy systems by neural networks is based upon the inherent capability of neural networks to perform this massive parallel processing of information. This is relevant to fuzzy controllers and more for fuzzy expert systems that processed large numbers of fuzzy inference rules in this real-time research.