intelligent systems ame 498g/598i. expert systems expert systems uses expert knowledge. expert...

28
Intelligent Systems AME 498G/598I

Post on 21-Dec-2015

224 views

Category:

Documents


2 download

TRANSCRIPT

Intelligent Systems

AME 498G/598I

Expert Systems

• Expert systems uses expert knowledge.• Expert knowledge is not limited to books.• It can also be privileged knowledge known only to

highly trained and experienced experts in a field.• Such knowledge is primarily intuitive.• Decisions are based on situation, judgement and

experience and it has a high probability of being correct.

• Computerized systems can be used as a database of experts decisions especially in engineering applications.

Database of Information

Rule Based Programming

Justify its Reasoning

Intelligent Advice/Decision

Expert Systems

• Rule based programming means the use of IF/THEN algorithmic structures to codify the decision rules from the expert.

• Engineers can be used to develop expert systems by using their expertise.

Areas of expertise of Expert Systems

• Synthesis• Engineering modeling• Taxonomy and morphology• Manufacturing methods• Maintenance• Applications• Tools• Optimization Criteria

Synthesis

• Engineers learns the laws of nature and procedures of analysis.

• Synthesis is the art of putting together a new machine, structure, chemical process, electrical circuit, or system.

• It is designer’s unique contribution to the art of engineering and the basis of industrial civilization.

• Duplicating synthesis in an expert system is far beyond our present knowledge of human thinking.

Engineering Modeling• Analysis is an essential component of design.• It ensures that the design is feasible and will meet the

performance requirements.• Analysis can also be used to examine that the design is

feasible in a optimum way.– Build engineering model of the physical model.

– Decide on the important failure modes.• Will a roof beam fail from excessive yielding of the material, or

collapse by buckling, or shearing of the end fasteners ?

• What level of snow loading should be applied ?

• Will a highway carry enough vehicles, have an adequate life, be safe, convenient, not require excessive maintenance, and permit snow removal under extreme conditions ?

Engineering Modeling

• Will an aircraft design have adequate life, reliability, capacity, controllability, and so on ?

• Will a transmission line carry the required current; or will an electrical motor provide the required power ?

• It requires engineering skill, experience, intuition and judgement for such predictions.

• Modeling of uncertainty is a major element of decision making for modeling.

• Poor judgement implies poor models.

Taxonomy and Morphology

• Engineering expertise is needed about the physical elements.

• New designs usually imply new combinations of existing elements.

• Advent of new element may not produce better designs than existing ones.– Proper knowledge about materials is a must.

• What is the best steel for both strength and corrosion resistance ?• What is the best lubricant for long life, minimum noise, minimum

friction loss, and minimum maintenance ?

Manufacturing Methods

• Engineer must also be an expert on manufacturing and building processes.– How metal parts are formed, cut or cast ?– How roads and buildings are constructed ?– How microchips are produced, and so on.

• Maintenance - knowledge of maintenance methods and testing procedures including special skills in diagnostic procedures and use of sensors.

• Applications – expertise is also needed on the field of application of the device or system being designed

• Tools – expertise on how to use the tools of the trade (using computers, calculators, etc)

Optimization Criteria• Expertise on the criteria to select the best possible

design.• Engineers normally work with immediate criteria.

– Maximize or minimize the performance criteria.• Minimum weight for an aircraft.• Minimum cost of a lathe.• Maximum capacity of a railway car, and so on.

• These can be achieved by analytical optimization or by judgement.

• Sometimes abstractions (deriving excitement, pleasure, aesthetics, etc) can be objectives and they are usually satisfied combining judgement and experience.

Super Expertise

• Aesthetics :– Ergonomics – human-machine interaction.– Styling.– Art of satisfying subtle human values.

• Risk judgement :– Subjective – Intuitive

Examples : Super Expertise

• Why did you use ductile iron rather than steel for that part ?– Ductile iron is cheaper than steel for a part that might be cast, and

it has excellent ductility for this service. Also it has worked well in previous similar applications.

• Why did you not heat treat that steel to give a yield strength of 250,000 psi (172 MPa) instead of only 125,000 psi (86.2 MPa) ?– I would like to have had the high yield strength that is possible with that

steel, but the ductility, impact strength, and resistance to crack growth are too low at that hardness. So I had to sacrifice yield strength to achieve them. The exact trade-off level that I chose is based on my experience in this application – It is intuitive judgement.

• Why did you use three bearings instead of two, in order to support that shaft in your gear box ?– We have found in this kind of application that only two bearings results in

excessive shaft deflection with corresponding misalignment of the gear teeth, and excessive wear and noise.

Expert Systems

• Experts express their knowledge in the form of rules for problem solving.– IF the ‘traffic light’ is green– THEN the action is go

– IF the ‘traffic light’ is red– THEN the action is stop

Rules for knowledge representation

• Rule is an IF-THEN structure that relates the information or facts in the IF part to some action in the THEN part. A rule provides some description of how to solve a problem.

• Any rule consist of two parts: the IF part, called the antecedent (premise or condition) and the THEN part called the consequent (conclusion or action).

• A rule can have multiple antecedents joined by the keywords AND (conjunction), OR (disjunction) or a combination of both.– IF <antecedent 1> IF <antecedent 1>

AND <antecedent 2> OR <antecedent 2>

AND <antecedent n> OR <antecedent n>

THEN <consequent> THEN <consequent>

Rules can represent relations, recommendations,directives, strategies and heuristics:

• Relation– IF the ‘fuel tank’ is empty

– THEN the car is dead

• Recommendation– IF the season is autumn

– AND the sky is cloudy

– AND the forecast is drizzle

– THEN the advice is ‘take an umbrella’

• Directives– IF the car is dead– AND the ‘fuel tank’ is empty– THEN the action is ‘refuel the tank’

• Strategy– IF the car is dead– THEN the action is ‘check the fuel tank step 1 is complete– IF the step 1 is complete– AND the ‘fuel tank’ is full– THEN the action is ‘check the battery’ step 2 is complete

• Heuristic– IF the spill is liquid– AND the ‘spill pH’ < 6– AND the ‘spill smell’ is vinegar– THEN the ‘spill material’ is ‘acetic

acid’

Development Team for Expert System

• Domain Expert

• Knowledge Engineer

• Programmer

• Project Manager

• End-user

Domain Expert

• Knowledge and skills for a specific domain.

• Expert has the greatest expertise in a given domain.

• His expertise is to be captured in the expert system.

• Expert should be able to communicate his or her knowledge, be willing to participate in the expert system development and commit a substantial amount of time to the project.

• The domain expert is the most important player in the development team.

Knowledge Engineer

• Capable of designing, building and testing an expert system.

• He interviews the domain expert to find out how a particular problem is solved.

• Finds out the reasoning methods the expert uses to handle facts and rules and decides how to represent them in the expert system.

• Decides on development software/expert system shell.

• Testing, revising and integrating the expert system into the workplace.

Programmer

• Responsible for actual programming of the expert system.

• Must be familiar with AI languages such as LISP, PROLOG, etc and the application of expert system shells.

• Must also know conventional programming languages such as C, Pascal, Fortran and Basic.

• Project Manager is the leader and coordinates the development of the expert system.

• End-user is the also called the user. Must be comfortable with the expert system. The inputs, outputs, etc.

LISPLISP does not evaluate expressions as a sequence of assignment statements.

define x, a and b y = ax + b z = y*y + sin(y) print z

Instead the computer is presented with a function, which it evaluates.

define x, a and b (f2 (f1 x a b))

f1 and f2 are functions as follows. f1 = (PLUS (TIMES a x) b) f2 = (PLUS (TIMES f1 f1) (SIN f1))

Algebraic type expressions are not used commonly.Functions are mostly operators that manipulate symbols.Symbols can be used to represent physical quantities or physical things and concepts.

(SETQ TOOLS (LIST : ‘HAMMER’, ‘SCREWDRIVER’, ‘SAW’, ‘WRENCH’)

Most widely used in artificial intelligence.

PROLOG

Started by Japanese.Higher level language than others.Specially designed for logic expressions

(IF/THEN).