Abhijit Gupta

A meta-heuristic approach to car allocation problem to reduce transportation cost over a fixed number of routes

This study addresses the problem of car allocation to different routes under certain restrictions with the objective of reducing the excess cost of transportation. A meta-heuristic approach based on Ant Colony Optimisation (ACO) algorithm is proposed and implemented to schedule the cars efficiently along the routes under the existing logistics. A mathematical model with two objectives is formulated for this purpose and solved in two phases. In the first phase, sequences of allocated cars are determined while in the second, car allocation scheme for each trip is determined using ACO algorithm. The simulation study is carried out from the empirical distributions of distance and time, followed by the sensitivity analysis on the basis of their stochastic behaviour. The cost benefit analysis shows a projected savings in terms of reduction of cost of travel, both with respect to distance and time, through the solutions obtained.

Optimization of product performance of a paint formulation using a mixture experiment

A paint manufacturing company was facing the problem of Vehicle Separation and Settling in one of its prime products. These two abnormalities are, in general, opposing in nature. The manufacturer tried several modifications in the existing recipe for the product but failed to control them. Experimentation was carried out using mixture design, a special type of designed experiment, and quadratic response surface models were fitted for both the responses. Finally, optimum formulation was obtained by simultaneously optimizing the two response surface models. During the determination of optimal formulation, different methods were compared. The optimum formulation is currently being used for regular manufacturing.

Generalized principal pivot transforms, complementarity theory and their applications in stochastic games

In this article, we revisit the concept of principal pivot transform and its generalization in the context of vertical linear complementarity problem. We study solution set and solution rays of a vertical linear complementarity problem. Finally we present an application of generalized principal pivot transform in game theory.

Online Control Charts for Process Averages Based on Repeated Median Filters

Two types of estimates of process level, namely repeated median estimates (Siegel, 1982) and full online estimates (Gather et al., 2006) based on repeated median filters, are used to develop control charts. The distributional properties of the estimates are studied using simulation and these are found to closely follow normal distribution. The repeated median being robust against outliers with asymptotically 50% breakdown value and having small standard deviation is found to be useful as a basis for monitoring process averages. The control charts using repeated median estimates have been recommended for general use.

ESTABLISHING OPTIMUM PROCESS LEVELS OF SUSPENDING AGENTS FOR A SUSPENSION PRODUCT

This article describes a study undertaken to find the optimum process levels of an antibiotic suspension product to be introduced in the market. Using an appropriate experimental design optimal process, levels were established. Confirmatory trials were ..