Fuzzy inference system or adaptive neuro-fuzzy inference system, and intelligent agent for the dynamic generation and retrieval of user interface software modules

A method of developing a fuzzy inference system (FIS) model of a target system for use in operating said target system, comprising the steps of:

• modelling a rate of change of an output vector of said target system as a product of a matrix of elements and a rate of change of an input vector of said system, elements of said matrix dependent upon antecedent, scaling and consequent parameters and at least one element from an input vector;
• receiving constraints on antecedent, scaling, and consequent parameters and input vectors, thereby defining sets of acceptable antecedent, scaling and consequent parameters and input vectors;
• receiving initial antecedent, scaling, and consequent parameters and an initial input vector;
• and
• minimizing a norm of said matrix to obtain optimized antecedent, scaling, and consequent parameters.

An intelligent agent for use in operating a user interface comprising:

• an active input interface for presenting user preference questions and receiving user preferences;
• an expert system for said user interface with initial membership functions;
• a fuzzy logic inference system (FIS);
• an adaptive neural network;
• and
• an output interface for controlling said user interface;
• said FIS and adaptive neural network for receiving constraints on antecedent, scaling, and consequent parameters and input vectors, thereby defining sets of acceptable antecedent, scaling, and consequently parameters and input vectors and for receiving initial antecedent, scaling and consequent parameters and an initial input vector from said active input interface and said expert system and for for each of a series of instants in time:
• modeling a rate of change of an output vector of said user interface as a product of a matrix of elements and a rate of change of an input vector of said system, elements of said matrix dependent upon antecedent, scaling, and consequent parameters and at least one element from an said input vector;
• and
• minimizing a norm of said matrix to obtain optimized antecedent, scaling, and consequent parameters and an optimized input vector.

A computer readable medium storing computer executable instructions that when loaded by a computing device, adapt said computing device to:

• provide an active input interface for presenting user preference questions and receiving user preferences;
• provide an expert system for said user interface with initial membership functions;
• provide a fuzzy logic inference system (FIS);
• provide an adaptive neural network;
• and
• provide an output interface for controlling said user interface;
• said FIS and adaptive neural network for receiving constraints on antecedent, scaling, and consequent parameters and input vectors, thereby defining sets of acceptable antecedent, scaling, and consequently parameters and input vectors and for receiving initial antecedent, scaling and consequent parameters and an initial input vector from said active input interface and said expert system and for for each of a series of instants in time:
• modeling a rate of change of an output vector of said user interface as a product of a matrix of elements and a rate of change of an input vector of said system, elements of said matrix dependent upon antecedent, scaling, and consequent parameters and at least one element from an said input vector;
• and
• minimizing a norm of said matrix to obtain optimized antecedent, scaling,, and consequent parameters and an optimized input vector.