Artificial Intelligence in Engineering 12 (1998) 121-134

0 1997 Elsevier Science Limited

Printed in Great Britain. All rights reserved
ELSEVIER PII: SO954-1810(97)00011-3 0954-1810/98/$19.00

Artificial intelligence approaches to determination

of CNC machining parameters in manufacturing:
a review
Kyung Sam Park & Soung Hie Kim*
Graduate School of Management, Korea Advanced Institute of Science and Technology, 207-43 Choengryangri, Dongdaemun,
Seoul, Korea

(Received 1 March 1995; in revised version 13 May 1996; accepted 5 February 1997)

In Computer Numerical Control (CNC) machining, determining optimum or

appropriate cutting parameters can minimize machining errors such as tool
breakage, tool deflection and tool wear, thus yielding a high productivity or
minimum cost. There have been a number of attempts to determine the
machining parameters through off-line adjustment or on-line adaptive control.
These attempts use many different kinds of techniques: CAD-based approaches,
Operations Research approaches, and Artificial Intelligence (AI) approaches.
After describing an overview of these approaches, we will focus on reviewing
AI-based techniques for providing a better understanding of these techniques in
machining control. AI-based methods fall into three categories: knowledge-based
expert systems approach, neural networks approach and probabilistic inference
approach. In particular, recent research interests mainly tend to develop on-line
or real-time expert systems for adapting machining parameters. The use of AI
techniques would be valuable for the purpose. 0 1997 Elsevier Science Limited.

Key words: CNC machining, machining parameter, knowledge-based expert

system, neural network, influence diagram.

1 INTRODUCTION Consequently, the conservative cutting conditions

Computer Numerical Control (CNC) machining is
widely used in mold/die industries and airframe
component manufacturing, because of its suitability
for high accuracy in machining complicated parts.’ In
the CNC machining, determining optimal cutting
conditions or parameters under the given machining
situation is difficult in practice. Conventional way for
selecting these conditions such as cutting speed and
feedrate, has been based upon data from machining
handbooks and/or on the experience and knowledge
on the part of programmer. The selected parameters,
in most cases, are extremely conservative to protect
excessive matching errors from tool failures such as tool
deflection, wear, breakage, etc. As a result, the metal
removal rate is low because of the use of such
conservative machining parameters.
As frequently encountered in complex or free-formed
surface machining, the geometry of the part or work-
piece prevents a constant depth and width of cut.
*Author to whom correspondence should be addressed.
127
assuming a constant depth and width of cut do not
perform high productivity. To overcome such a prob-
lem, the machining parameters should be adjusted
according to the current in-process part geometry.
The objective of this paper is to review prior work on
determining machining parameters in order to give a
better understanding to researchers and practitioners
in machining domain, since we have not found any
publication on the survey. In particular, we will focus on
examining Artificial Intelligence (AI)-based methods. In
the next section, we describe an overview of related
work that attempts to select an optimal machining
parameters, and present AI-based approaches in the
subsequent sections.
2 AN OVERVIEW
Recently, there have been several attempts to deter-
mine the optimal machining parameters from off-line
adjustment or on-line adaptive control, thus the part
programmer does not have to spend time and effort to
128 Kyung Sam Park, Soung Hie Kim
calculate their optimal values. These attempts are
categorized as CAD-based approaches, Operations
Research (OR) approaches, and AI approaches.‘3
2.1 CAD-based approaches
The off-line approach uses machining process models,
cutting force and tool wear models, based on a prior
knowledge gathered from off-line experiments. Based on
the process models, cutting force and tool wear are
calculated through computer machining simulation
(CMS) using information on NC-code with initial
machining parameters, tool shape and workpiece
geometry. Using the results, an optimum machining
parameter for each tool motion is achieved by maxi-
mizing the metal removal rate (MRR) without violating
machining constraints.
The basic concept of optimizing machining param-
eters is that when the cutting force is too large at the
large depth and width of cut, either low feedrate or high
cutting speed, or both can be added to the NC-code.
However, note that too high cutting speed can not be
selected since the tool life is largely due to the cutting
speed.5 Most CAD-based approaches belong to the off-
line adjustment. Advantages of these methods are that
they are easy and effective in practical applications.
Figure 1 shows a conceptual framework for simulation
and optimization of machining.
Machining process models
The machining process models represent the rela-
tionships between the machining responses (i.e., cutting
force and tool wear) and the machining conditions in a
specific tool and workpiece. These models can be built
by prior knowledge obtained from field and laboratory
experiments. An example of machining process model
based on a multiplicative model is given by
Cutting Force (N), FC = ald2f a3dn4waS,
Tool Life (min.), TL = blvb2f b3db4wb5,
where V, f, d and w are, respectively, cutting speed
(mm/min), feedrate (mm/tooth), depth of cut (mm),
and width of cut (mm); and oi and J3i are the model
parameters.
( /,+
Computer Machining Simulation
Fig. 1. A framework for simulation
The multiplicative model can be generated from
statistically planned machining tests (see Ref. 11). An
advantage of this model is easily obtained even though
geometry of cutting tool is complicated such as ball-end
mill. Furthermore, it is reported the accuracy of the
models is quite good. ‘J~J’ The above models will play a
role of constraints in optimizing machining parameters.
In addition, analytical process models for the prediction
of cutting force have been studied,7~8110V17 and their
application to on-line feedrate adjustment in end milling
has been found in Ref. 6. However, a difficulty may exist
in using such analytic models in practice because of their
high computational complexity.
Computer machining simulation
The main objective of CMS for determining machining
parameters is to compute the maximum depth of cut (d)
and width of cut (w) for each tool motion from given
part geometry, NC-code and tool configuration. Why
compute the maximum point? The reason is machining
error from the tool failures is mostly occurred at the
maximum point. CMS of in-process workpiece can be
realized as a Boolean subtraction of the space occupied
by the tool movement along the tool path from initial
part geometry. Hence, it is first needed to represent the
part geometry for CMS.
Solid modeling’8>‘912’or Z-buffer techniques22>23have
been used to model workpiece geometry for CMS. A
pape?’ has proposed a method of feedrate adjustment
using a swept volume generation technique based on
solid modeling. However, this method can calculate
only average cutting force, thus it does not provide the
instantaneous cutting force that is necessary for
estimating the tool failures. Z-buffer model is a
form of discrete nonparametric representation in
which the Z-values of the surface are given at grid
points on the XY-plane. More detailed description
on Z-buffer model and its application to control
and monitoring of machining can be found in the
literature.22’23
Machining parameter optimization
Based on the CMS and the machining process models,
feedrate (f ) and cutting speed (v) are determined in the
chining Process Models I
and optimizaiton of CNC machining.
 Determination of CNC machining parameters
optimization module. For increasing the productivity,
MRR has to be maximized while maintaining an
allowable load fluctuation on the cutting tool in spite
of variations in depth of cut and width of cut. The MRR
is expressed as MRR = kvfdw, where k = n/(rD), n is
the number of tooth, and D the diameter of the
tool.’ A mathematical model for such problem can be
formulated as follows:
Model 1: Maximize MRR = kvfdw
subject to v,in 5 v 5 vmax
FC I JG,,,
TLmin I TL 5 TL,,,
HP 5 HPr,,,,,
where HP represents the spindle horsepower
(Nmm/min) as a constraint for the machine capacity,
and I’min and Vmax, respectively, are minimum and
maximum allowable values of V. HP is expressed based
on the FC as HP = c - FC - v, where c is 0.041 as a
constant.5 Once taking natural logarithms in Model 1,
it is converted into the standard linear programming
(LP) form. Thus, the LP problem can be solved by
using a general algorithm referred to as the Simplex
method.24
2.2 Operations research approaches
Of course, the use of the above LP technique can be
viewed as an OR approach. However, main research
interest of OR approaches is to minimize global
machining cost by considering multiple criteria related
to machining, thus which problem is to solve a multiple
criteria optimization problem (for an overview of the
multiple criteria optimization problem, see Refs 25,
26). These methods should be used for off-line
adjustment because of the restriction of computational
time. An advantage of these methods can provide a
reference model, i.e., a general model because an
exhaustive consideration on selecting machining
parameters is involved. A typical research is found in
Ref. 27. According to the research, the model without
describing full mathematical form can be expressed as
follows:
Model 2: Maximize {MMR}
Minimize {surface roughness}
Minimize {machining cost}
subject to the constraints of Model 1.
On surface roughness, there are two methods of its
measuring: root-to-crest roughness and roughness
average.2y3 The factors of measuring machining cost
per workpiece are cost of tool, cost of cutting, and costs
associated with machine idle time, due to setup, loading
129
and unloading and tool changes to replace worn-out or
damaged tools. Solving Model 2 is more complex
because the model have multiple objectives and con-
flicting between the objectives (a mathematical repre-
sentation and the solution method for Model 2 appear
in Ref. 27).
2.3 Artificial intelligence approaches
The on-line approach is an attempt to automatically
adapt and optimize the machining parameters based on
sensor information on machining responses in real time,
without CMS. Well-known sensor information is listed
as cutting force, tool wear, tool temperature and
acoustic emission. Note that the information of tem-
perature and acoustic emission can not be used in off-
line methods using CMS. These information, however,
can play very important role in machining control
or adapting machining parameters, praticularly when
occurring an abnormal machining due to unpredictable
variables such as unknown material properties, tool
conditions, etc.
For on-line control, the following components or
techniques are required: (1) sensing devises, (2) repre-
senting the information from the sensor, and (3)
optimizing machining parameters. A description of
sensing devises is not presented in this paper. For the
description, refer to Refs 4, 16.
AI approaches offer a possible technique in order to
handle the problems (2) and (3). One of the most
important factors for successive on-line control is the
execution time with respect to machining control or
determining optimal machining parameters. Reaction to
machining conditions by tool wear, machine break-
downs and other failures must be carried out within
seconds or milliseconds to guarantee the safety and
reliability of the machining process. However, a
simplistic adaptation of AI techniques to machining
control would be inadequate, because execution time of
these systems are generally too long as compared with the
reaction time required for the machining control,
particularly if the knowledge base becomes very complex.
There have been a number of studies on the
application of AI techniques to on-line control, which
we categorize into knowledge-based expert systems
approach, neural networks approach and probabilistic
inference approach. Each approach is described in the
next subsequent sections.
3 KNOWLEDGE-BASED EXPERT SYSTEMS
APPROACH
3.1 An overview
Knowledge-Based Expert Systems (KBES) are intel-
ligent computer programs that capture the specific
130 Kyung Sam Park, Soung Hie Kim
knowledge of a particular domain and mimic the
problem-solving strategies of human experts to provide
recommendations?8-3o They represent a new problem-
solving paradigm that utilizes many techniques developed
from AI research. The KBES can capture causal and
inferential knowledge about machining processes to
provide expert-level recommendations during decision-
making processes and hence are valuable aids to
machining operators who face increasingly complex tasks.
With the KBES technique, machining control deci-
sions using the sensor information can be made to
maintain the machining parameters within critical
constraints. Strictly speaking, on-line control with the
KBES is an adaptive control of satisfying machining
constraints, simply stated ACC, rather than an adaptive
control with optimization, ACO. Whereas AC0 systems
seek to adjust machining parameters in a direction that
optimize a predefined performance index, i.e., objective
such as MRR, the aim of ACC systems is that the
machining parameters are adjusted to their maximum
possible values given the constraints of the machining
process.’ Recent research on machining control using
KBES techniques has been found in Refs 31, 32. In the
next subsection, we will describe a KBES framework for
machining control and present an example of simple
production rules for the determination of machining
parameters.
3.2 Structure of KBES for adaptive control
A structure of the KBES approach for machining
control is shown in Fig. 2. It consists of three modules:
a knowledge base, an inference engine, and a sensor data
acquisition and processing module. The inference engine
drives the system and interfaces with the knowledge
base and hence supplies advice to the user and an
explanation to justify the system’s line of reasoning. The
knowledge base can provide near-optimal machining
control with experimental data. The methods for
inference can be modeled as rules, e.g., IF (antecedent)
THEN (consequence).
To achieve the near-optimal machining parameters
and machining control such as tool change, machining
Knowledge Base
Fig. 2. A KBES structure
stop, and so on, it is necessary that an adaptive control
algorithm that uses the recursive adaptive model and
the constraint rules is developed. For example, the
constraint rules can be expressed as shown in Fig. 3.
In KBES, many techniques for knowledge representa-
tion have been developed, for instance, production
rules, semantic nets, frames, etc. The type of knowledge
representation that is appropriate in a given situation
depends on what sort of knowledge is being represented
and how it is to be applied. In time-critical machining
control applications, it is imperative the knowledge
representation scheme is efficient. Among the KBES
approaches to machining control, in Ref. 31 a frame-
based scheme is used, and in Ref. 32 a production rule
representation is applied such as shown in the above
paragraph.
4 NEURAL NETWORKS APPROACH
4.1 An overview
Neural networks differ in various ways from con-
ventional expert systems to traditional computing. The
reasons are as follows. First, unlike traditional expert
systems where knowledge is made explicit, neural nets
generate their own knowledge by learning from domain
examples. This means that neural nets can easily
make the knowledge base by learning, and they do not
require additional knowledge acquisition processes
which require enormous time and efforts in the expert
systems. Supervised learning is achieved through the
learning rule which adapts the connection weights of the
network in response to the inputs and the desired output
pairs. Many other network learning rules have been
inverted also in Ref. 34.
Second, neural computing is both distributed and
associative in knowledge representation.33 The dis-
tributed and associative nature of neural net leads to
a reasonable response even when presented with
incomplete or previously unseen input. In particular,
multi-layer neural nets which register in their hidden
layers important features of the knowledge domain,
for machining control.
 Determination of CNC machining parameters 131

feedforward neural net with one hidden layer is shown

Rulesfor Finding MachiningSituation: in Fig. 4. Each node or processing element (PE) in every
layer is fully connected to other PE in the proceeding
layer, and ever PE sums its weighted inputs and passes
through some kind of transfer function such as linear
or sigmoid functions. The learning parameters are the
Rule 1: IF (NOT (A&c,,?,,, ‘: Acstic c AcsticJ)
connection weights and the PE’s parameters, i.e.,
/* Acstic = Intensity of Acoustic etmss~?n from sensor */
threshold values.
THEN (Stop the machining operatmn AND Change tool If necessary) As a learning rule, Rumelhart and McClelland33 have
Rule 2: IF (NOT (Temp nil” < Temp - Temp,> ,,,”)) developed the generalized delta rule called back-
I* Temp : Temperature from sensor *I
propagation algorithm that is basically a gradient
method. This rule aims at minimizing the global error
THEN (Stop the machining operation AND Change tool if necessary)
of the system by adjusting the learning parameters. The
backpropagation algorithm does not always find global
Rule 5: IF ((FC[k] J FC[k]J AND (feedm” c feed))
minimum but may stop at a local minimum. However,
/* FC[k]-Cuttmg Force at axis k. k=x,y, from sensor. feed=Feedrate ‘1 in most cases, the system can usually be driven to the
THEN (feed = feed 0 01 until FC[k] L FC[kJ,J global minimum or to the desired accuracy with an
Rule 6: IF ((FC[k] -z FC[k],J AND (t‘eed I feed”,,,\) AND (speed speed,,,J)
appropriate choice of hidden PEs. The number of
THEN (speed = speed + 0 02 until FC[k] < FC’[k],,,,,b)
hidden PEs must be large enough to form a decision
region that is as complex as required by the given
Rule 7: IF ((FC[k] -’ F(‘[k],,,,\) AND (feed < feed,,,J)
problem, and on the other hand is small enough that the
THEN (feed = feed + 0 01 unti! FC[k] : FC[k]“J
generalization ability remains good.

Rule n: IF ((FC[k] :’ FC[k],J AND (feed 1 feed”,,,\) AND
THEN (Stop machinmg operatmn)
Fig. 3. An example of rules for adapting machining parameters.
can use this hidden knowledge to generate non-trivial
generalizations.
In machining domain, neural nets can possess abilities
to learn from experience and to use the knowledge
gathered during the learning process to optimize the
machining control. Experience is represented by input-
output data, where input variables are machining
parameters such as feedrate and cutting speed and
output variables are signals from sensors such as cutting
force, tool wear, temperature and acoustic emission. The
aim of learning is to establish a generalized mapping
between the input and output, where note that this
statement is a supervised learning. This section deals
with a supervised learning approach (see Refs 35,36). For
unsupervised learning approaches refer to Refs 37, 38.
In addition, there are several techniques for repre-
senting the input-output relationship: multiple
regression (see Section 2.1), the group method of data
(speed ? speed,,>J) 4.3 Optimal control phase
Assuming that a neural network has been trained by the
procedure mentioned previously, then the objective of
optimal control phase is to determine an appropriate
machining parameter that optimizes a performance
index, given machining constraints on the network
outputs.
Let us consider a trained neural net with n input PEs
and m output PEs. Let ai and dip respectively, be actual
and desired output of the ith output PE. For k of the m
output nodes, the di represent the desired outputs of the
PEs, whereas for the remaining m - k PEs, the di is the
maximum allowable outputs of the PEs (e.g., cutting
force and horsepower). Then, a performance index PI
is defined by PI = w1 - ERR - w2. MRR, where MRR
represents metal removal rate (see Section 2.1)
ERR = 5 (di - ai)2/2,
i=l
output
layer

handling, and the neural network (see next subsection).

A study4 has reported that the neural network is the
most effective method for tool wear identification Hidden
layer
through a comparative analysis among the above
techniques. Furthermore, neural net knowledge-base
acquired during the learning phase can subsequently be
Input
used for determining optimal machining parameters. layer

4.2 Learning phase

Fig. 4. A neural net structure for representing machining
For building machining knowledge base, a two-layer knowledge.
132 Kyung Sam Park, Soung Hie Kim
and w1 and w2 are constants that represent the relative
importance of ERR and MRR.
Machining optimization problem is to find the n
inputs, denoted by pi, that minimize PI subject to the
following constraints:
Pmini I Pi I Pmaqr i= 1 ,...,n,
di - ai > 0, i=k+l,...,m.
The solution method of this constrained minimization
problem can be found in Ref. 35.
5 PROBABILISTIC INFERENCE APPROACH
Agogino et a1.43 have proposed an influence diagram
as a framework for integrating machining operator’s
expertise, first-principle knowledge and experimental
data for the wide range of sensors possible for in-process
monitoring and control. The use of multiple sensors
reduces the sensitivity of the system to any specific
sensor’s drawbacks. The non-deterministic or prob-
abilistic nature of the inference problem and noisy
sensor data is handled by operations with Bayesian
probability.
Influence diagram has been developed for represent-
ing complex decision problems based on incomplete and
uncertain information from a variety of sources.39~44
Knowledge of the interrelationships between variables is
represented in a compact graphical and numerical
framework which identifies the critical variables and
explicitly reveals any conditional independence between
them.
The knowledge representation using influence dia-
grams can be viewed from three hierarchical levels:
topological, functional and numerical level.39 At the
topological or relational level, the nodes in the diagram
represent the key variables in the system being modeled,
and the arcs or arrows identifies conditional influences
or functional relations between the nodes. In the CNC
machining, examples of key variables are machining
parameters and machining responses from sensor
Fig. 5. An influence diagram
information. The nature of the influences is specified
at the functional level and further quantified at the
numerical level.
From the discussion in the preceding two paragraphs,
influence diagram is defined by an acyclic directed-graph
G = (N, A) with A c N x N: It contains three types of
nodes in the node set N. The chance node, which is
circular shape, represents uncertain or certain states
(e.g., cutting force, tool wear, acoustic emission), the
rectangular-shaped decision node (e.g., feedrate, cutting
speed) reveals a variables whose value is chosen by the
decision maker, and the diamond-shaped value node
(e.g., metal removal rate) represents the objective to be
maximized in expectation by the decision analysis. It
should be noted that influence diagrams on the
topological level do not need a mathematical or
probabilistic basis to justify themselves. Their influences
are justified by mathematical or probabilistic repre-
sentation at the functional level. At the final level,
numerical level, utilities of the decision maker, and
probability distributions from prior information by
experiments are assessed numerically for each node.
Shown in Fig. 5 is a simple example of an influence
diagram for machining optimization.
Once a complete influence diagram is generated, the
diagram is manipulated and evaluated for determining
the optimal decision strategy. A direct solution proce-
dure to automate influence diagrams has been proposed
in Refs 40-42. This algorithm consists of the value-
preserving translations, node removal and arc reversals,
which correspond to the rollback procedure in deci-
sion tree models.45 For more detailed description on
applied influence diagrams to machining monitoring
and control, see Ref. 43.
6 CONCLUDING REMARKS
This paper presented a survey of prior studies on
determining an optimal machining parameter and
machining control. We particularly focused on reviewing
for determining machining parameters.
 Determination of CNC machining parameters
AI-based methods for on-line adaptive control, i.e.,
KBES, neural networks and probabilistic inference
approaches. These approaches are commonly based on
the sensor information of machining responses in real
time, and based upon the prior knowledge from
machining experiments in field and laboratory, such
as machining constraints, machinability data (i.e.,
mechanical and material data of cutting tool and
workpiece), etc. The prior knowledge can be elicited
from data of machining handbooks (e.g., see Ref. 14)
or the experience and knowledge of the part
programmer.
Observe that the difference that underlies AI
approaches is only the style of knowledge base being
represented in the system and its reasoning (or inference)
for machining optimization and control. That is: The
KBES approach uses a knowledge base such as
production rules (see Ref. 32) and frame-based tech-
niques (see Ref. 31), with an inference engine. The
neural networks approach uses a feedforward network
model (see Refs 35, 36) based on supervised learning
rules or self-organization network model (see Refs 37,
38) based on unsupervised or competitive learning
rules, without an additional inference engine. The prob-
abilistic approach uses an influence diagram model (see
Ref. 43) with a probabilistic reasoning engine.
In building an intelligent machining control system, a
principle problem is how to represent knowledge in a
way that is suitable for a particular machining domain.
There have not been a specified guideline for the
choice of a knowledge representation scheme. The
general or conceptual guidelines are efficiency and
sufficiency, where sufficiency means that the precision
of the predictions we obtain meets our requirements,
and efficiency refers to their practical applicability. All
AI approaches mentioned in this paper may be sufficient
for representing a machining knowledge, although their
knowledge representation styles are different. If the
above statement is true, the most important criteria
becomes efficiency.
There may be some factors for measuring efficiency in
quantitative or qualitative manner, for example, easy for
naive user and reaction time. Among them, reaction
time would be the most important factor in real-time
machining control systems. Namely, reaction to
machining failures and unfavorable machining has to
be made within seconds or milliseconds to guarantee the
safety and reliability of the machining process. In this
point of view, we prefer a knowledge representation
scheme in which reaction time is less sensitive as the size
of the knowledge base increases.
Apart from the sufhciency, what kind of knowledge
representation technique is better than the others in a
time-critical machining control system? This paper can
not give an answer for this question, because there
have not been a comparative analysis among the AI
approaches. The problem in order to address this
question is a promising further research issue.
133
1. Balakrishnan, P. and Dervries, M. F., Sequential estima-
tion of machinability parameters for adaptive optimization
of machinability data base systems. ASME Journal of
Engineering for Industry, 1985, 107, 159- 166.
2. Bhattacharya, A. R., Faria-Gonzalez and Ham, 1. Y.,
Regression analysis for predicting surface finish and its
application in the determination of optimum machining
conditions. ASME Journal of Engineering for Industry,
1970,92,711-714.
3. Chandiramani, K. L. and Cook, N. H., Investigation on
the nature of surface finish and its variation with cutting
speed. ASME Jounal of Engineering for Industry, 1964,86,
134-140.
4. Chryssolouris, G. and Domroese, M., An empirical study
of strategies for integrating sensor information in
machining. Annals of CZRP, 1989, St?(l), 425-428.
5. DeGarmo, E. P., Black, J. T. and Kohser, R. A., Materials
and Processes in Manufacturing. MacMillan Publishing
Co., New York, 1984, pp. 429-629.
6. Fusell, B. K. and Srinivasan, K., An investigation of the
end milling process under varying machining conditions.
ASME Journal of Engineering for Industry, 1989, 111,
27-36.
7. Kline, W. A., DeVor, R. E. and Lindberg, J. R., The
prediction of cutting forces in end milling with application
to cornering cut. International Journal of Machine Tool
Design and Research, 1982, 22(l), 7-22.
8. Kline, W. A., DeVor, R. E. and Shareef, I. E., The
prediction of surface accuracy in end milling. ASME
Journal of Engineering for Zndustry, 1982, 104,272-278.
9. Koren, Y., Computer Control of Manufacturing Systems.
McGraw-Hill Co., New York, 1983, pp. 193-220.
10. Martellotti, M. E., An analysis of the milling process.
Transactions of the American Society of Mechanical
Engineers, 1941, 63, 677-700.
11. Neter, J., Wasserman, W. and Kutner, M. H., Applied
Linear Statistical Models: Regression, Analysis of Variance,
and Experimental Design. Irwin Inc., Homewood, Illinois,
1985.
12. Park, K. S., Analysis of Cutting Force in Ball-End Milling
Process (in Korean). M.E. thesis, Dept. of Industrial
Engineering, Korea Advanced Institute of Science and
Technology, Korea, 1991.
13. Park, K. S. and Kim, S. H., Use of AI-based approaches
for determining CNC machining parameters in manu-
facturing. In Encyclopedia of Microcomputers, eds A. Kent
and J. G. Williams. Marcel Dekker, Inc., New York,
1995.
14. Thompson, R. W., Tipnis, V. A. and Kegg, R. L. (eds),
Machinability Testing and Utilization of Machining Data.
Materials/Metalworking Technology Series, American
Society for Metals, Metals Park, Ohio, 1979.
15. Tipnis, V. A., Buescher, S. C. and Garrison, R. C.,
Mathematically modeled machining data for adaptive
control of end milling operations. Proceedings of the
Fourth NAMRC, 1976, 279-286.
16. Tlusty, J. and Andrews, G., A critical review of sensors for
unmanded machining, Annals of the CZRP, 1983, 32(2),
563-572.
17. Tlusty, J. and MacNeil, P., Dynamics of cutting forces in
end milling. Annals of the CZRP, 1975, 24(l).
18. Mantyla, M. An Introduction to Solid Modeling, Computer
Science Press Inc., 1988.
19. Wang, W. P. and Wang, K. K., Geometric modeling for
swept volume of moving solids. ZEEE Computer Graphics
and Applications, 1986, 6, 8- 17.
20. Wang, W. P., Solid modeling for optimizing metal
 Kyung Sam Park, Soung Hie Kim

removal of three-dimensional NC end milling. Journal of optimization for intelligent machining. Journal of
Manufacturing Systems, 1988, 7(l), 57-65. Manufacturing Systems, 1991, 10(6), 464475.
21. Hook, T. V., Real-time shaded NC milling display. 33. Rumelhart, D. and McClelland, J., Parallel Distributed
Computer Graphics, 1986, 20(4), 15-20. Processing, Vol. 1. MIT Press, Cambridge, 1986.
22. Jerard, R. B. and Drysdale, R. L., Geometric simulation 34. Lippmann, R., An introduction to computing with neural
of numerical control machining. Proceedings of ASME nets. IEEE Transactions on ASSP, 1987, 4(2), 4-22.
International Computers in Engineering Conference, 1988, 35. Rangwala, S. S. and Dornfeld, D. A., Learning and
129-136. optimization of machining operations using computing
23. Takata, S., Tsai, M. D., Inui, M. and Sata, T., A cutting abilities of neural networks. IEEE Transactions on
simulation system for machinability evaluation using a Systems, Man, and Cybernetics, 1989, 19(2), 299-314.
workpiece model. Annuls of the CZRP, 1989, 38(l), 36. Dornfeld, D. A., Neural network sensor fusion for tool
417-420. condition monitoring. Annuls of CZRP, 1990, 39(l),
24. Taha, H. A., An Introduction to Operations Research. 101-105.
MacMillan Publishing Co., New York, 1982, 15-157. 37. Burke, L. I., Competitive learning based approaches to
25. Chankong, V. and Haimes, Y. Y., Multiobjective Decision tool-wear identification. IEEE Transactions on Systems,
Making: Theory and Methodology. North Holland, New Man, and Cybernetics, 1992, 22(3), 559-563.
York, 1983. 38. Burke, L. I., An unsupervised approach to tool wear
26. Steuer, R. E., Multiple Criteria Optimization: Theory, identification. ZIE Transactions, 1993.
Computation, and Application. John Wiley & Sons, New 39. Howard, R. A. and Matheson, J. E., Influence diagrams.
York, 1986. In The Principles and Applications of Decision Analysis,
27. Malakooti, B. and Deviprasad, J., An interactive multiple Vol. 2, eds R. A. Howard and J. E. Matheson. Menlo
criteria approach for parameter selection in metal cutting. Park, CA, Strategic Decision Group, 1984.
Operations Research, 37(5), 805-8 18. 40. Shachter, R., Evaluating influence diagrams. Operations
28. Hayes-Roth, F., Waterman, D. A. and Lenat, T. J., Research, 1986, 34, 871-882.
Building Expert Systems. Addison-Wesley, Reading MA, 41. Shachter, R., Probabilistic inference and influence
1983. diagrams. Operations Research, 1988, 36, 589-604.
29. Buchanan, B. G. and Shortliffe, E. H., Rule-Based Expert 42. Rege, A. and Agogino, A. M., Topological framework for
Systems: The MYCZN Experiments of the Stanford representing and solving probabilistic inference problems
Heuristic Programming Project. Addison-Wesley, Reading in expert systems. IEEE Transactions on Systems, Man,
MA, 1984. and Cybernetics, 1988, 18(3), 402-414.
30. Lu, S. C-Y. and Komanduri, R. (eds), Knowledge-Based 43. Agogino, A. M., Srinivas, S. and Schneider, K. M.,
Expert Systems for Manufacturing. The American Society Multiple sensor expert system for diagnostic reasoning,
of Mechanical Engineers, New York, 1986. monitoring and control of mechanical systems. Mechanical
31. Lingarkar, R., Liu, L., Elbestawi, M. A. and Sinha, N. K., Systems and Signal Processing, 1988, 2(2), 165-185.
Knowledge-based adaptive computer control in 44. Oliver, R. M. and Smith, J. Q. (eds), Influence Diagrams,
manufacturing systems: a case study. IEEE Belief Nets and Decision Analysis. John Wiley & Sons, New
Transactions on Systems, Man, and Cybernetics, 1990, York, 1990.
20(3), 606-618. 45. Bunn, D. W., Applied Decision Analysis. McGraw-Hill
32. Billatos, S. B. and Tseng, P. C., Knowledge-based Inc., New York, 1984.