A Deterministic Inventory/Production Model with General Inventory Cost Rate Function and Concave Production Costs

Ş. İlker Birbil, Hans Frenk, Pelin BayindirRePub 2004

Abstract

We present a thorough analysis of the economic order quantity model with shortages under a general inventory cost rate function and concave production costs. By using some standard results from convex analysis, we show that the model exhibits a composite concave-convex structure. Consequently, an effective solution procedure, particularly useful for an approximation scheme, is proposed. A computational study is appended to illustrate the performance of the proposed solution procedure.

Introduction: Beyond Linear Costs

Traditional inventory theory, epitomized by the Economic Order Quantity (EOQ) and Economic Production Quantity (EPQ) models, often relies on the simplifying assumption of linear costs. However, real-world industrial processes frequently exhibit economies of scale, where producing larger batches reduces the marginal cost per unit, leading to concave production costs. Similarly, inventory holding and backorder penalty costs are rarely perfectly linear.

This paper addresses a deterministic production model that integrates these complexities: concave variable production costs alongside general, non-linear inventory holding and backorder penalty functions. The objective is to determine the optimal "up-to" inventory level ($S$) and cycle time ($T$) that minimize the total average cost per unit time.

Inventory level over time in the generalized production model.

Methodology: Exploiting Mathematical Structure

The authors leverage convex analysis to navigate the optimization challenges posed by non-convex cost functions. Their approach involves several critical steps:

  1. Structural Reformulation: The problem is shown to be separable. By defining an auxiliary function for the optimal inventory cost over a single cycle, the authors demonstrate that the overall problem possesses a "C-programming" structure—a composition of concave and convex functions.
  2. Transformation to Convexity: A pivotal mathematical insight involves a change of variables. By replacing the cycle time $T$ with its reciprocal $1/T$, the inner optimization problem is transformed into a convex one, making it globally solvable.
  3. Piecewise Linear Approximation: To maintain high modeling flexibility for "general" cost functions while ensuring mathematical tractability, the authors employ piecewise linear segments to approximate non-linear curves.
Piecewise linear approximation of a concave cost function.
  1. Bisection Algorithm: In the piecewise linear case, analytical expressions are derived that facilitate the use of a bisection method to find the optimal cycle time. This method is chosen for its exponential convergence rate.

Key Insights and Results

The discovery of the hidden concave-convex structure is the paper's most significant contribution. It allows for global optimization in a domain typically fraught with local minima. By explicitly modeling concave production costs, the framework captures the strategic shift toward larger batch sizes inherent in economies of scale.

Computational experiments using randomly generated problem instances demonstrated the algorithm's efficiency. Implemented in MATLAB on a 1 GHz processor, the average solution time for complex problems (e.g., 20 breakpoints in the inventory cost function and 10 in the production function) was a mere 0.18 seconds. For typical scenarios, the optimal policy was identified in under 0.05 seconds, indicating high suitability for real-time Enterprise Resource Planning (ERP) systems.

Limitations and Future Directions

While robust, the model assumes deterministic (constant and known) demand. Future research could extend this framework to stochastic environments, which would better reflect market volatility. Additionally, the current model focuses on single-item optimization; extending it to multi-item scenarios with shared production capacities or shipping constraints remains a valuable area for further study. Finally, while piecewise linear approximations are effective, establishing formal error bounds relative to the original non-linear functions would enhance the model's theoretical rigor.

Summary updated

Abstract

We present a thorough analysis of the economic order quantity model with shortages under a general inventory cost rate function and concave production costs. By using some standard results from convex analysis, we show that the model exhibits a composite concave-convex structure. Consequently, an effective solution procedure, particularly useful for an approximation scheme, is proposed. A computational study is appended to illustrate the performance of the proposed solution procedure.
BibTeX
@inproceedings{birbil2004deterministic,
  author = {Ş. İlker Birbil and Hans Frenk and Pelin Bayindir},
  title = {{A Deterministic Inventory/Production Model with General Inventory Cost Rate Function and Concave Production Costs}},
  booktitle = {RePub},
  year = {2004},
}
Figure