A COMPUTATIONAL COMPARISON OF LINEAR OPTIMIZATION SOLVERS IN PYTHON FOR AGGREGATE PRODUCTION PLANNING MODELS

Abstract

Aggregate Production Planning (APP) plays a dynamic role in leveling production rates, manpower levels, and inventory over a planning scope based on demand prediction. This paper investigates a comparative analysis of three open-source Python solvers PuLP(using the CBC solver), SciPy(using the linprog method), and CVXPY(using the SCS solver) to solve the APP problem under a fixed workforce model in textile industry. Instead of introducing productivity loss as in some earlier models, this research keep constant workforce levels to simplify manpower-related cost and to emphasize solver efficiency. The first goal is to measure the computational efficiency, solution accuracy, and utility of each solver when applied to the same linear programming(LP) formulation of the APP problem. On a constant data set, all solvers are tested to assure fairness in comparison. The results highlight the capability and limitations of each solver, offering valuable modality for researchers and professionals intent to select proper tools for production planning project using Python-based optimization.