Page 22 - IJOCTA-15-1

S. Karahan Orak, N. Aydin, E. Karatas / IJOCTA, Vol.15, No.1, pp.14-24 (2025)

economic growth increases demand or crises decrease de- approach. Various studies have been conducted on sto-
mand are examples of this. Similarly, the product life cycle, chastic modeling in different fields in subsequent years.
which includes stages such as growth, maturity, decline, C¸imen et al. 15 recommended establishing appropriate plan-
and death, causes changes in current demand. Therefore, ning models and solution methods for managing inventory
non-stationary demand is frequently encountered, and the systems under uncertainty, as we did with stochastic mod-
dynamic demand structure is crucial for many goods. 8 eling for inventory planning. C¸eki¸c 10, working in the
same discipline, proposed stochastic programming-based
Sample Average Approximation (SAA) is a sample-based solutions for the applicability of inventory problems with
approach to solving stochastic optimization problems in- non-stationary demand and fixed order costs in multi-
volving large scenario numbers and complex uncertainty stock point systems, presenting stochastic programming as
structures. SAA provides computational efficiency for such a flexible modeling approach for this challenging inventory
problems, as it can handle large scenario numbers more ef- control problem. Focusing on a different aspect, Toksarı
fectively and offer faster solution times. Although they do et al. 32 worked on optimizing multi-stage supply chain
not exactly match the true optimal solution, the results management by considering uncertainties in defect rates
can be considered approximate optimal solutions. Ad- between procurement, production, and distribution stages,
ditionally, iterative improvement and application flexibil- highlighting the concept of defect.
ity are useful in various industrial scenarios. In terms of
cost-effectiveness, it is less costly than full stochastic solu- Several solution methods were presented for representing
tion methods. For these reasons, SAA is a common solu- stochastic modeling problems, and Monte Carlo Simula-
tion method in stochastic programming problems involving tion (MCS) is one of them. Sarit et al. 33 addressed the
large data sets. 14, 23, 24 challenges related to inventory management of products
with stochastic demand. Their study used MCS to create
In factories or manufacturing plants, production is carried
distributions of demand and delivery times for inventory
out by applying several methods, The quantity of products
items. It was then used to estimate potential profit and risk
to be produced or can yield maximum profit at minimum
associated with inventory policies. Many studies are also
cost can be estimated using stochastic models. Multiple
available in the literature on chance-constrained stochas-
methods can be used together when generations use sto-
tic programming as a different type of solution approach
chastic parameters. Scenario-based forecasting is one of 34
for stochastic modeling. Atalay et al. examined the
them. These scenarios represent how the system will work,
equivalence of deterministic models to chance-constrained
and the results will be produced under specified conditions.
stochastic programming models when the coefficients of
The results found are important for comparing the poten-
random variables follow normal and chi-square distribu-
tial outcomes arising from the decision variables. Different
tions, contributing to this field. Similarly, Aksaraylı et
disciplines have conducted studies in the field of industrial al. 2 used chance-constrained stochastic programming to
pipe clamp manufacturing. Wiseman et al. 25 conducted a plan the production of an office supplies manufacturing
study on the design of dynamic pipe clamps. Malakov et company.
al. 26 conducted a study on selecting the optimal size range
for pipe clamps. This study applies SAA to solve various scenarios based on
different sample sizes. Zhang et al., 35 using different sam-
Ozturk 27 conducted a study focused on enabling domestic ple sizes in their study, developed a method for stochastic
companies to compete with imported products by opti- programs with complementary constraints where equilib-
mizing energy efficiency and brass pipe fittings produc- rium constraints can be replaced with smooth functions.
tion quantities. Similarly, Ka¸ctıoglu and Oz¨ukl¨u 28 ap- Bastin et al. 36 applied a variable sample size technique
plied fuzzy goal programming to maximize profits while to estimate choice probabilities in solving unconstrained
calculating the optimal production quantities for a firm mixed logit models. Byrd et al. 37 developed a method-
manufacturing various metal components. G¨uls¨un et al. 29 ology based on variable sample size to solve large-scale
developed an integrated production planning model for a machine-learning problems.
clamp manufacturing company in their study. To solve the
model, they employed a Preference-Based Optimization This article aims to achieve a result that evaluates the
method known as Linear Physical Programming (LPP), stochastic situation affecting constraints under uncertain
and they obtained results aimed at minimizing costs and demand and optimizes the objective function using SAA,
reducing the impact of hiring and firing decisions on work- which supports solving a large sample set.
force motivation. Mukherjee and Ray 30 critically analyzed
various modeling and optimization techniques in metal- The study’s contribution may be as follows: The com-
cutting processes. They proposed parameter optimization mon point in studies examining stochastic modeling but
strategies to ensure the advantages of appropriate selec- addressing different solution approaches is the desire to
tion. Nystrom and Soderstrom 31 improved processes in handle random demand. In this study, we performed
a multi-facility metal industry company, increasing pro- scenario-based production planning for a factory producing
ductivity. Through their standardized work practices, industrial-type pipe clamps with stochastic modeling. To
they demonstrated that combining these tasks resulted in the authors’ best knowledge, no study in the literature ad-
greater profitability. Various studies in the literature focus dresses optimal production planning for pipe clamps. This
on the design, quality, and strength of pipe clamps across study examines the production planning of this important
different disciplines. component in the industry for the first time using sto-
chastic modeling. Moreover, as inventory planning mod-
Stochastic modeling was introduced to the literature by els are generally designed deterministically, this study con-
George B. Dantzig in 1955 with his work ”Linear Pro- tributes to the literature with its stochastic modeling ap-
gramming under Uncertainty,” the first example of this proach. Additionally, various constraints and parameters
16