Fuzzy multi-objective supplier selection considering risk and a strategic plan,

Magdalena Kalata, Dobrila Petrovic

Research output: Contribution to conferenceAbstractpeer-review


The paper concerns a selection of suppliers in a large real-world manufacturing supply network. The suppliers are selected considering three objectives including minimization of the total cost and risk and maximization of the strategic plan achievement. Two decision variables are introduced to handle risky suppliers - safety time for suppliers with consistent delivery problems and safety stock for critical products supplied by suppliers with a high non-conformance rate. The risk is modelled using two parameters: the product risk (related to the product complexity and the authority of the relevant supplier) and the supplier risk (related to the quality and delivery performance indicators). The achievement of the strategic plan considers categorisation of the suppliers into four categories: E-exit suppliers with bad scores, Mmaintain suppliers with good scores difficult to replace, G-grow suppliers with increasing importance and N-new strategically important suppliers. Uncertain contributions of the four supplier categories to the achievement of the strategic plan are described by imprecise terms such as low, medium, high or very high and are modelled using fuzzy numbers. Depending on user preferences, each objective can be assigned different importance. The fuzzy multi-objective optimisation model is transformed into a corresponding crisp optimisation model with a single objective to maximise satisfaction degree of attaining the optimal value of each objective.
Original languageEnglish
Number of pages1
Publication statusPublished - 2016
Event28th European Conference on Operational Research - Poznan, Poland
Duration: 3 Jul 20166 Jul 2016


Conference28th European Conference on Operational Research
Internet address

Fingerprint Dive into the research topics of 'Fuzzy multi-objective supplier selection considering risk and a strategic plan,'. Together they form a unique fingerprint.

Cite this