Title:Modeling the Workpiece Roughness Using Fuzzy Logic Modeling
Volume: 6
Issue: 3
Author(s): Hussien M. Al-Wedyan
Affiliation:
Keywords:
ANFIS, cutting parameters, first order Sugeno, fuzzy logic, lathe machine, roughness.
Abstract: In this paper, first order Sugeno-fuzzy logic-based models are constructed using the cutting parameters as an
input data and the surface roughness (roundness) as an output data. Mapping between the input-output spaces is constructed
to find the effect of the input cutting parameters on the output surface quality in terms of roundness error which is
an indication of the surface roughness. Hence, the best cutting conditions in the lathe machine are selected to improve the
output. A method is proposed to accurately establish the relationship between the different cutting parameters and the resulting
surface irregularities which form the surface roughness using subtractive clustering technique based first order
Sugeno fuzzy model. Model having minimum errors with fewer numbers of rules for each surface irregularity is obtained
through enumerative search of the clustering parameters. The resulted model with optimum clustering parameters is then
tuned by using adaptive neuro-fuzzy inference system (ANFIS). Consequently, the proposed model is successfully used to
predict and minimize the roughness error using different cutting parameters involved in this investigation. Consequently,
the machinist will decide beforehand which tool he should use before conducting the experiment, which will save time, effort
and money. Hence, the best parameters can be picked up to achieve a reduced roundness error. The first model considers
two cutting parameters as an input to the process. The second model will include tool usage as a third input. It
shows that the tool usage should be included due to its effect in deteriorating the surface quality. In the second model,
minimum roundness equals to 0.998 m with 21.1 Hz, 0.0944 mm/rev, and 3.08 for workpiece frequency, feed rate and
tool usage, respectively. The purpose of this study is to provide an overview about the use of first order Sugeno-fuzzy
logic-based models in machining processes by addressing recent patents and scholarly articles on the subject of design,
control and application of such systems.
