Expert Systems and Adaptive Control Robert Stengel Robotics and Intelligent Systems MAE 345 Princeton University 2013 Expert systems Gain scheduling Adaptive critic Cerebellar model articulation controller Reinforcement learning Failure tolerant control Copyright 2013 by Robert Stengel All rights reserved For educational use only http www princeton edu stengel MAE345 html Expert Systems Using Signals to Make Decisions Programs that exhibit intelligent behavior Program that uses rules to evaluate information Program meant to emulate an expert or group of experts making decisions in a specific domain of knowledge or universe of discourse Program that chains algorithms to derive conclusions from evidence Functions of Expert Systems Design Diagnosis Negotiation Impart knowledge or skill Explain or analyze observations Monitoring Observe a process compare actual with expected observations and indicate system status Propose assess and prioritize agreements between parties Planning Prediction Devise actions to achieve goals Reason about time forecast the future Reconfiguration Interpretation Determine the nature or cause of an observed condition Instruction Conceive the form and substance of a new device object system or procedure Alter system structure to maintain or improve performance Regulation Respond to commands and adjust control parameters to maintain stability and performance Principal Elements of a RuleBased Expert System Critical Issues for Expert System Development System architecture Inference or reasoning method Deduction Knowledge acquisition Induction Explanation Abduction User interface Representation of Knowledge for Inference Logic Predicate calculus 1storder logic Fuzzy logic Bayesian belief network Search Given one state examine all possible alternative states Procedures Function specific routines executed within a rigid structure e g flow chart Semantic propositional networks Model of associative memory Tree or graph structure Nodes objects concepts and events Links interrelations between nodes Production rule based systems Rules Data Inference engine Basic Rule Structure Rule sets values of action parameters Rule tests values of premise parameters Forward chaining Reasoning from premises to actions Data driven facts to conclusions Backward chaining Reasoning from actions to premises Goal driven find facts that support a hypothesis Analogous to numerical inversion Elements of a Parameter Type Name Current value Rules that test the parameter Rules that set the parameter Allowable values of the parameter Description of parameter for explanation Elements of a Rule Type Name Status 0 1 T F U Has not been tested Being tested Premise is true Premise is false Premise is unknown Parameters tested by rule Parameters set by rule Premise Logical statement of proposition or predicates Action Logical consequence of premise being true Description of premise and action for explanation The Basic Rule IF THEN ELSE If A TRUE then B else C Material equivalence of propositional calculus extended to predicate calculus and 1st order logic i e applied to logical statements Methods of inference lead to plans of action Compound rule Logic embedded in The Basic Rule e g Rule 1 If A B and C D then perform action E else Rule 2 If A B or C D then E F else Nested pre formed compound rule Rule embedded in The Basic Rule e g Rule 3 If A B then If C D then E F else else Finding Decision Rules in Data Identification of key attributes and outcomes Taxonomies developed by experts First principles of science and mathematics Trial and error Probability theory and fuzzy logic Simulation and empirical results Example of On Line Code Modification Execute a decision tree Get wrong answer Add logic to distinguish between right and wrong cases If Comfort Zone Water then Animal Hippo else Animal Rhino True but Animal is Dinosaur not Hippo Ask user for right answer Ask user for a rule that distinguishes between right and wrong answer If Animal is extinct Decision Rules Representation of Data Set Crisp sets Fuzzy sets Schema Diagrammatic representation A pattern that represents elements or objects their attributes or properties and relationships between different elements Frame Hierarchical data structure with inheritance Slots Function specific cells for data Scripts frame like structures that represent a sequence of events Database Spreadsheets tables graphs Linked spreadsheets Structure of a Frame Structure array in MATLAB Structure or property list in LISP Object in C Ordered set of computer words that characterize a parameter or rule An archetype or prototype Object oriented programming Express Rules and Parameters as Frames Example Fillers and Instance of a Frame Application Specific Frame Generic Fillers Instantiation Legal fillers Can be specified by Inheritance and Hierarchy of Frame Attributes Data type Function Range Inheritance property All instances of a specific frame may share certain properties or classes of properties Hierarchical property Frames of frames may be legal Inference engine Decodes frames Establishes inheritance and hierarchy Executes logical statements Animal Decision Tree Forward Chaining What animal is it Premise Rule 1 Parameter S i z e If Small Else test Action Parameter None Premise Rule 2 END Parameter S o u n d If Squeak Animal Else A n i m a l Action Parameter Animal Premise Rule 3 END Premise Rule 5 END END Giraffe Elephant Comfort Parameter Comfort Zone If Water Animal Else A n i m a l Action Parameter Animal Mouse Trunk Parameter T r u n k If True Animal Else test Action Parameter Animal Squirrel Parameter N e c k If Long Animal Else test Action Parameter Animal Premise Rule 4 END test Sound Neck Rhino Zone Hippo END Animal Decision Tree Backward Chaining What are an animal s attributes Animal Hippo From Rule 5 Comfort Zone From Rule 4 Trunk False From Rule 3 Neck Short From Rule 1 Size Large Water Animal Decision Tree Parameters Type Object Attribute Type Object Attribute Name A n i m a l Name N e c k Current Value Variable Current Value Variable Rules that Test None Rules that Test 3 Rules that Set 2 3 4 5 Allowable Values Mouse Squirrel Giraffe Elephant Rules that Set None Allowable Values Long Short Hippo Rhino Description Neck of Animal Description Type of Animal Type Object Attribute Name S i z e Current Value Variable Rules that Test 1 Rules that Set None Allowable Values Large Small Description Size of Animal Type Object Attribute Name T r u n k Current Value