Knowledge Based System For Design
Knowledge Based System For Design
Presented edit Master subtitle style Click to by Sujit M. Mulay (P11CC013) Guided by Dr. S. Kumar
2/16/13
an knowledge based system for design/selection of bending die components, Following objectives are identified to fulfill the said goal.
1. 2. 3. 4. 5.
2 2
Literature Review. Identification of major die components. To Prepare Production Rules based on Knowledge acquired from various experts. To develop modules for major components of bending die. Selection of suitable programming language.
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED 2/16/13 6. Execution of the SY3STEM system. OF BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
Outline
Introduction To Bending Die. Bending Die Components. Need of Expert System In Die Design. Expert System Methodology. Development Procedure of Expert System for die design. Literature Review. Tentative Research Plan for Next Semester. References.
3 3
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 OF BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
axis.
Bending is done using press brakes. Capacity of press brake varies from 20 to 200tons.
4 4
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 OF BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
parts have already replaced many expensive cast, forged, and machined products.
Bending is one of oldest forming process. It is a flexible process. It is a mass production process. Variety of different shapes can be produced. It is through metal bending that the process of creating a
5 5
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
Bending die
A bending die is a specialized tool used in manufacturing
6 6
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 OF BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
V bending die
7 7
KNOWLEDGE BASED SY3STEM 2/16/13 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF FOR DESIGN/SELECTION OF BENDING DIE COMPONENTS
U bending die
8 8
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
9 9
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
10 10
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF BENDING KNOWLEDGE BASED SY3STEM 2/16/13 DIE COMPONENTS FOR DESIGN/SELECTION OF
component that houses the opening and receives punches. Material Tool Steel
Punch plate Similar to die block used to hold and
on
gives actual shape to the part. Material - hardened steel or tungsten carbide.
Guide post It function together with guide bushings to
Material Chemical Silicon Steel (S50C), Grey Cast Iron (FC250), Stainless Steel (SS400).
12 12 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 OF BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
It is a heavy plate, ribbed with T slots precision aligned to the frame with dowel pins. Material - steel
Cushion pin - Metal pins used in conjunction with a die
cushion to transfer pressure from the cushion to the bottom of a die pad, It can also be used to lift the part out of the die. Material Cold roll steel
13 13 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 OF BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
design and manufacturing is largely depends upon the skill and the experience of the die designer.
Poor Documentation of die designing rules. Commercially available CAD/CAM software's are less
efficient. They provide some assistance in drafting and analysis in die design process, but human expertise is still need to arrive at the final design. time and cost of product.
14 14
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
Expert System
Definition Expert system = knowledge + problem-solving methods Expert systems (ES) is a branch of applied artificial
intelligence involving that expertise, which is the vast body of task-specific knowledge, is transferred from a human to a computer. This knowledge is then stored in the computer and users call upon the computer for specific advice as needed (Liao, 2005)
An expert system is a system that employs human
knowledge captured in a computer to solve problems that ordinarily require human expertise (Turban,1995)
KNOWLEDGE BASED SY3STEM 2/16/13 15 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION FOR DESIGN/SELECTION OF 15OF BENDING DIE COMPONENTS
techniques.
recognized as a practical tool for solving real-world problems. Universities offered expert system courses and two thirds of the Fortune 1000 companies applied the technology in daily business activities.
16 16
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 OF BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
modied, if necessary.
Cost: The per-user cost is greatly lowered. Permanence: will last indefinitely (depending on the hardware). Reliability: provide a second opinion to human experts, mediate
opinions.
Explanation: the expert system may always explain how it reached the
conclusion.
Response: may provide fast or real-time response for critical
applications.
17 17 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 OF BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
18 18
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
Expert system Methodology Determination of goal & preliminary investigation Problem analysis, System draft, System prototype & validation of the prototype In house testing using a real life case study Complete testing of real life data of user System validation and documentation Fixing of error and continuous enhancements
19 19 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF BENDING KNOWLEDGE BASED SY3STEM 2/16/13 DIE COMPONENTS FOR DESIGN/SELECTION OF
- System is being designed. The needed degree of integration with other subsystems and databases is established. - Concepts that best represent the domain knowledge are worked out. - Inferencing should be established with sample cases to represent the knowledge and to perform.
20 20 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 OF BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
Development of Prototype knowledge engineer works with the expert to place the
initial kernel of knowledge in the knowledge base. specific tool chosen for the project
- Using sample cases, the prototype is tested, and deficiencies in performance are noted. End users test the prototypes of the ES.
21 21
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
Interaction of the ES with all elements of its environment, including users and other information systems, is ensured and tested. ES is documented and user training is conducted
- System is keep current primarily by updating its knowledge base. - Interfaces with other information systems have to be maintained as well, as those systems evolve.
KNOWLEDGE BASED SY3STEM 2/16/13 22 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION BENDING DIE COMPONENTS 22 OF FOR DESIGN/SELECTION OF
23 23
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 OF BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
24 24
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF BENDING KNOWLEDGE BASED SY3STEM 2/16/13 DIE COMPONENTS FOR DESIGN/SELECTION OF
design of bending die is collected through on line and off line consultation with design experts, tool design engineers of different industries, referring research articles, catalogue and manual of different design and manufacturing industries.
Knowledge Representation -
Production rule based - The most common method of knowledge representation is ruled based systems. The knowledge collected is represented using rules. The syntax of a production rule is IF < condition >, THEN < action >
25 25 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
rules framed for each module crosschecked from die design expert by presenting them IF condition of the production rule of IF THEN variety. The framed rules are presented in structured manner. hardware elements depending upon memory requirement, processing speed and needed configuration should be selected. The proposed systems are implemented on PC (Pentium 4 CPU, 2.4 GHz, 2GB of RAM) with Autodesk AutoCAD 2008.
far in AI systems are FORTRAN, KEE, OPS, PROLOG, TURBOPROLOG and LISP.
26 26
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF BENDING DIE KNOWLEDGE BASED SY3STEM 2/16/13 COMPONENTS FOR DESIGN/SELECTION OF
systems realized very early that by separating the knowledge base containing the problem specific information from the rest of the system.
Most of that system could be easily adapted to a new
problem by replenishing the knowledge base with information about the new problem.
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 27 BENDING DIE COMPONENTS 27 OF FOR DESIGN/SELECTION OF
shells, offering much of the functionality of an expert system through the user interface, dialog component, inference engine, and explanation generator while leaving the population of the knowledge base to the application programmer.
knowledge base of the system are linked together by an inference mechanism, which makes use of forward chaining.
forward chaining Data driven process. Backward chaining - Goal driven process.
To enables the user to input the essential sheet metal component data. To displays the optimal decision choices for the users benefit.
28 28
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
Literature Review
Sr. Researcher No
1 Uzsoy, et. al., (1991)
System Details
Rule based KBC implemented using turbo prolog.
Remarks
A simple sheet part with total of eight features; four holes with two holes radius of 2 units, two hoes of radius 1 unit and four bends. System does not provide process planning and manufacturability of bending Data base for present m/c required to be made
2 Ching, Z. et.al. (1994) 3 Ching, Z. et. al. (1996) 4 Ong, S. et al. (1997) 5 Gupta, et al., (1998)
Autolisp integrated into AutoCAD Sheet metal bending machine selection model
Fuzzy set theory for sheet metal Does not provide optimal bending sequences bend sequencing only provides feasible solution Automated process planning for Separate illustrations are given for each of the robotic sheet metal bending modules for different operations; press brake Presently sheet with 23 bends can be addressed for its process planning & other sequences.
29 29
KNOWLEDGE BASED SY3STEM 2/16/13 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF FOR DESIGN/SELECTION OF BENDING DIE COMPONENTS
Sr. No Researcher
6 Inamdar, M. et al. (2000) Ching, Z. (2001) Shigeru, A. et al. (2002)
System Details
Artificial Neural Network for measuring Sprinback in v-Bending Torque equilibrium and optimal strip working sequence for bending progressive die Sheet metal Bending Sequence and robot grasping positions are determined by Graphical method and geometrical bending features
Remarks
More the Data for initial training the more accurate will the result Two stage model generated for customized problem Only in 55% cases model provides accurate result and around 27 % moderate result.
Ruffini, R. (2002)
Neural network for springback minimizationInitial data is required for training of ANN model so, not suitable for initial use where precise o/p is required CAPP process planning of Bending and Piercing System required commercial CAD/CAM software
10
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION KNOWLEDGE BASED SY3STEM 2/16/13 30 FOR DESIGN/SELECTION OF 30 OF BENDING DIE COMPONENTS
Sr. No
11
Researcher
Rico, J.C. et al. (2003) Rico, J.C. et al. (2003) Ehrismann, R et al.(2004) Pathak, K. et al. (2005) Sousa, L.C. et al.(2006)
System Details
Remarks
Automatic bending Sequence for Model is based on Part-tool collision and Parallel bends tolerance constraint with lower process time optimisation Automatic bending Sequence for Model is based on Part-tool collision and Parallel bends tolerance constraint with lower process time optimisation Expert system for Bending Sequence System is not able provide optimized solution
12
13
14
Neural Network for finding sheet To reduce the error % of results in b/w metal bending process FEM and NNT high no. of cases are parameters required for training Numerical simulation coupled Time required for numerical simulation is with Genetic Algorithms used to High around 36 h. optimized V & U bending processes
15
16
Ching, Z. et Selection of bending Good for customized problem al.(1998) tools only and bending sequence
KNOWLEDGE BASED SY3STEM 2/16/13 31 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION BENDING DIE COMPONENTS 31 OFFOR DESIGN/SELECTION OF
Sr. No
17
Researcher
Bozdemir, M. et al. (2008) Kazan, R. et al. (2009) Kontolatis, N. et al. (2010)
System Details
ANN for Springback in V-Bending ANN for springback in wipe bending Optimization of Bending Process planning parameters
Remarks
Good results are obtained for customized problems Optimized results are not obtained System is not capable to provide solution for sprinback in present approach Experimental data is used for learning model
18
19
20
KNOWLEDGE BASED SY3STEM 2/16/13 32 KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION FOR DESIGN/SELECTION OF 32OF BENDING DIE COMPONENTS
References
Boljanovic, V., Sheet Metal Forming Processes and Die Design, Industrial Press New
York, 2004.
Suchy, I., Handbook of Die Design, McGraw Hill Publication, 2006. Donaldson, C., Tool Design, Tata McGraw Hill Publication, 1976. Groover, M.P., Fundamentals of Modern Manufacturing, Materials Processes, and
1987.
Stamping Die Design Textbook, Jetro Supporting Industry Program, 2006. Gowari, S., Hariharan, P., Manufacturing Technology 1, Dorling Kindersley (India)
Pvt.ltd. 2008.
Parik, S.B. et.al., A CAD/CAM system for deep drawing dies in a simple-action press,
Journal of Materials Processing Technology Volume 87, Issues 13, 15 March 1999, Pages 258265
F., M., "Press Brake Bending: Methods and Challenges", Metal forming: 3843. (August
34 34
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
2008).
edition, 2009
Krishnamoorthy, C. S., Rajeev, S., Artificial Intelligence and Expert Systems for
Analytical Tools, Narosa Publishing House, New Delhi, India and Springer Verlag, New York, 1992.
Tisza, M., Development of Expert System for Process Planning of Deep Drawing,
Applications: A Survey, journal of Intelligent and Robotic Systems 10, pp. 113-145, 1994.
Minsky, M. L., A Framework for Representing Knowledge in The Psychology of
35 35
Buchman, B. G., Shorliffe, E. H., Rule based Expert system, Addison-Wensley, KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 Reading, Mass, 1984 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
Frost, R., Introduction to knowledge based systems, Collins Professional and Technical
Books, 1986
Kowalski, R. A., logic for problem solving, North Holland, Amsterdam, 1979 Kocabs, S., a review of learning, the knowledge engineering review 6(3) 1991 Zone-Ching, L., Da-Yuan, C., Selection of sheet metal bending machines by the PRISM-
in the selection system of sheet metal bending tooling, Artificial Intelligence in Engineering 10, pp. 21-37, 1996.
Zone-Ching, L., Dar-Yuan, C., The Selection System for Sheet Bending Tooling, Int. J
sequencing for sheet metal bending ,Journal of Materials Processing Technology 69, pp. 29-36,1997.
Gupta, S. K., Bourne, D. A., Kim, K.H., Krishnan, S.S., Automated Process Planning for
Sheet Metal Bending Operations. Journal of Manufacturing Systems Vol. 17/No. 5, 1998.
36 36
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
Ruffini, R., Using Neural Network for Springback Minimization in a Channel Forming
bending of metallic sheets using an artificial neural network, Journal of Materials Processing Technology 108, pp. 45-54, 2000.
Baseri, H.M., Rahmani, B., Modeling of spring-back in V-die bending process by using
fuzzy learning back-propagation algorithm, Expert Systems with Applications 38, pp. 88948900,2011.
Kontolatis, N. G., Vosniakos, C., Optimisation of press-brake bending operations in 3D
sheet metal using neural network, Materials and Design 30, pp. 418423, 2009.
37 37
KNOWLEDGE BASED SY3STEM FOR DESIGN/SELECTION OF KNOWLEDGE BASED SY3STEM 2/16/13 BENDING DIE COMPONENTS FOR DESIGN/SELECTION OF
Thank
You.
2/16/13