Multi-Objective Supplier Allocation Model

We consider the least restrictive case where any of the buyers can acquire one or more products from any suppliers, namely, a multiple sourcing model. 

In this phase of the supplier selection process an organization will make the following decisions  

The mathematical model for the order allocation problem is discussed next. 

---

For the critical items in Quadrant 3, we present a three-step method for global supplier selection. In the first step, we present a GP model for country selection this step shortlists a country using various qualitative and quantitative criteria. In the second step, we assess the risks of supply using the analytic hierarchy process (AHP). In the final step, we develop a multi-objective model with price and risk as the two conflicting objectives. For every product, we assign three different suppliers—a global supplier, a domestic primary supplier, and a domestic secondary supplier. Order allocation among the suppliers is optimally decided by the model. 

This chapter presents a disruption risk quantification method and a multiobjective supplier selection model to generate mitigation plans against disruption risks. The proposed risk quantification method considers risk as a function of two components—impact and occurrence. Impact is modeled using GEVD distributions, and occurrence is assumed to be Poisson-distributed. The disruption risk quantification method calculates the estimated value of the loss due to disruptive events at a supplier, which is then used in a multi-objective optimization model. The model minimizes cost, lead time, and risk and then maximizes quality and determines the optimal supplier and order allocation for multiple products. The model is solved using four different GP solution techniques—preemptive, non-preemptive, min-max, and fuzzy GP Optimal solutions are displayed using the VPA, and the performance of the solution techniques is discussed. We observe that, for the data set we have tested, preemptive GP, non-preemptive GP, and min-max GP achieve three out of four objectives. 

---