Sie sind auf Seite 1von 27

Expert system-based optimization in wrought copper alloy production

Dr.-Ing. George Power, Universidad de Lima gepower@correo.ulima.edu.pe

Contents
Introduction to expert systems Rule-based expert system for job preparation in wrought copper alloy production Programming components Results obtained Summary and conclusions

02/11/2006

Copyright George Power

What is an expert system?


Software for specific problem areas involving expert knowledge Developed after AIs failure of General Problem Solver Commercial XPS can be:
Case-based Rule-based Decision trees

02/11/2006

Copyright George Power

Case-based expert systems


Databases of specific problems and their solutions For a given problem, the expert system tries to find a similar case and its solution Example: Medical diagnosis appropriate treatment for a patient with concrete symptoms

02/11/2006

Copyright George Power

Rule-based expert systems


Have a large set of if-then-else rules, algorithms, formulas, databases, etc. Business Rule Management Systems (BRMS) are specific for business aplications Rules represent general cases from which solutions are drawn Rules must be entered and maintained by experts
02/11/2006 Copyright George Power 5

Decision trees
Expert systems for classification tasks Inductive learning based on a set of examples Object attributes are knots of a search path or tree, characteristics represent links Goal is the leaf or end knot gives the object class to be assigned

02/11/2006

Copyright George Power

Individual roles in expert systems


End-user
Interacts with the XPS and can also supply some expertise

Problem domain expert


Builds and supplies the knowledge base, can also be the programmer

Knowledge (software) engineer


Assists the expert in representing his knowledge
02/11/2006 Copyright George Power 7

Semis of wrought copper alloys

Source: www.wieland.de
02/11/2006 Copyright George Power 8

Manufacturing of flat products

Source: www.wieland.de
02/11/2006 Copyright George Power 9

Some fabrication parameters for Cu alloy sheet and strip


Material description (alloy type, temper, standards, ) Other technological properties (flexural, die wear, soldering tests) Hardening and softening behaviour (cold-work and annealing curves) Surface finish, protection Fabrication steps: hot rolling, cold rolling, annealing, pickling, stretching, slitting, traverse winding, etc.
02/11/2006 Copyright George Power 10

Why an expert system for job preparation?


Time-consuming consults to experts Huge amount of data in production and customer specific procedures Delays in job preparation increase lead times in PPS Order changes and cancellations (storno) Reduce overhead costs
02/11/2006 Copyright George Power 11

Team structure
Department Job preparation Programming tasks Object structure, production steps, resources (machines, materials), input and output forms Machine setup, production, maintenance and down time, cost factors Material properties, machine specifications, databases, formulas
12

Time and cost calculation Quality Engineer

02/11/2006

Copyright George Power

enginObject: An object-oriented programming tool for XPS


Development system
Shop specific planing know-how Planing documents Planing algorithms

User system
CAD PPS

Knowledge base

Create plan
CAM SML Job plans SQL

Materials Machines Tools Decision tables tn = Formulas Databases

Preparation Set up Soldering Texty blocks

Edit / actualize job plans

Source: camos GmbH, 1994

02/11/2006

Copyright George Power

13

enginObject structure
Components:
Decision tables Procedures Databases Forms 2D tables Formulas Methods Objects Variables

Development system:
Class browser Compiler Interpreter / Debugger

User system:
Class browser Run-time Interpreter

02/11/2006

Copyright George Power

14

Decision tables
Allow precise modelling of complex logic structures in a compact format Clear table structure replaces several nested if-then-else clauses Conditions, rules and actions are coded as entered Standardized in DIN 66241 (1979): Decision tables for information exchange A special case of decision tables are twodimensional (2D) tables
02/11/2006 Copyright George Power 15

Decision tables

Rules Conditions

Actions

02/11/2006

Copyright George Power

16

More examples

Decision table with priorization of machine selection

2D table for rolling cost factor assignment.


02/11/2006 Copyright George Power 17

Procedures
Allow semi-structured text programming of sentences, loops, arrays, etc. Syntax similar to Basic or Pascal without strict variable declaration Calls to other components (decision tables, formulas, other procedures) within the same object

02/11/2006

Copyright George Power

18

Procedures
Logical decision Call decision table Call formula Comment line String constant Numeric constant

Reserved word Variable (variant type)

02/11/2006

Copyright George Power

19

Databases
Allow access to structured data (material properties, standard tempers, etc.) Simple query structures Advanced queries with SQL server for accessing external databases

02/11/2006

Copyright George Power

20

Databases (example: material properties)

02/11/2006

Copyright George Power

21

Forms (input/output)

02/11/2006

Copyright George Power

22

Class browser (object tree)

Colapsed tree structure

Expanded tree structure

02/11/2006

Copyright George Power

23

Interpreter and debugger


Useful for test runs during program development Once tested, a run-time version of the expert system can be compiled File extension is .knb (knowledge base) Users access expert system from a local or remote server through the run-time interpreter.
02/11/2006 Copyright George Power 24

Interpreter

02/11/2006

Copyright George Power

25

Results
System programmed and maintaned by same experts. Realization time ca. 8 months 10+ years in use, extended to other production sites running remotely on same Citrix server Reduction of personnel in job preparation, employes can be reoriented to other activities Reduced lead times Increased product quality More precision in cost calculation XPS never forgets a single rule or procedure less quality claims Succesful experience in job preparation applied for another expert system: technical assistance of sales department.
02/11/2006 Copyright George Power 26

Conclusions
Expert systems allow efficient management of one of the companys main assets: technical know-how Programming tools are easy to learn. Expert and programmer tasks can be performed by the same personnel within the company, freeing IT resources Knowledge base can be accesed remotely and must be maintained regularly by the same programming team Limitations in structured programming arent a drawback: they simplify documentation and maintenance.
02/11/2006 Copyright George Power 27

Das könnte Ihnen auch gefallen