Optimization of a dynamic supply portfolio considering risks and discount’s constraints
Abstract: Nowadays finding
reliable suppliers in the global supply chains has become so important for
success, because reliable suppliers would lead to a reliable supply and besides
that orders of customer are met effectively . Yet, there is little empirical
evidence to support this view, hence the purpose of this paper is to fill this
need by considering risk in order to find the optimum supply portfolio.
Design/methodology/approach: This paper proposes a multi objective model
for the supplier selection portfolio problem that uses conditional value at
risk (CVaR) criteria to control the risks of delayed, disrupted and defected
supplies via scenario analysis. Also we consider discount’s constraints which
are common assumptions in supplier selection problems. The proposed approach is
capable of determining the optimal supply portfolio by calculating
value-at-risk and minimizing conditional value-at-risk. In this study the
Reservation Level driven Tchebycheff Procedure (RLTP) which is one of the
reference point methods, is used to solve small size of our model through
coding in GAMS. As our model is NP-hard; a meta-heuristic approach,
Non-dominated Sorting Genetic Algorithm (NSGA) which is one of the most
efficient methods for optimizing multi objective models, is applied to solve
large scales of our model.
Findings and Originality/value: In order to find a dynamic supply
portfolio, we developed a Mixed Integer Linear Programming (MILP) model which
contains two objectives. One objective minimizes the cost and the other
minimizes the risks of delayed, disrupted and defected supplies. CVaR is used
as the risk controlling method which emphases on low-probability,
high-consequence events. Discount option as a common offer from suppliers is
also implanted in the proposed model. Our findings show that the proposed model
can help in optimization of a dynamic supplier selection portfolio with
controlling the corresponding risks for large scales of real word problems.
Practical implications: To approve the capability of our model various
numerical examples are made and non-dominated solutions are generated.
Sensitive analysis is made for determination of the most important factors. The
results shows that how a dynamic supply portfolio would disperse the allocation
of orders among the suppliers combined with the allocation of orders among the
planning periods, in order to hedge against the risks of delayed, disrupted and
defected supplies.
Originality/value: This paper provides a novel multi objective model for
supplier selection portfolio problem that is capable of controlling delayed,
disrupted and defected supplies via scenario analysis. Also discounts, as an
option offered from suppliers, are embedded in the model. Due to the large size
of the real problems in the field of supplier selection portfolio a
meta-heuristic method, NSGA II, is presented for solving the multi objective
model. The chromosome represented for the proposed solving methodology is
unique and is another contribution of this paper which showed to be adaptive
with the essence of supplier selection portfolio problem.
Keywords: Supplier selection,
Dynamic supply portfolio, Conditional value-at-risk, Mixed integer programming,
RLTP, NSGA II
Author: Masoud Rabbani, S.M
Khalili, H Janani, M Shiripour
Journal Code: jptindustrigg140032
