| Autor: |
Dasgupta, Arup, Mahapatra, Amalendu Singha, Santra, Prasun Kumar, Mahapatra, Ghanshaym Singha, Shaw, Ashok Kumar, Sarkar, Biswajit |
| Předmět: |
|
| Zdroj: |
Journal of Industrial & Management Optimization; Nov2024, Vol. 20 Issue 11, p1-38, 38p |
| Abstrakt: |
In supply systems, maintaining stock levels is a major challenge. The purpose of this article is to determine the number of orders in a continuous review model. The demand rate is erratic, as it is dependent on promotional advertising on an ambiguous basis. This article considers learning and memory in an unpredictable setting. The fundamental crisp model is extended with a fuzzy formulation, and learning is included. Demand promotion is included with the complete backlog in models, which are based on finite-horizon time scales. The memory present in real-world situations is represented using Caputo's style of nonintegral ordered calculus. It focuses on the percentage of each process that is devoid of scarcity as well as the overall cycles of replenishment. After analyzing each model numerically, a parametric sensitivity analysis is performed to show the trade-offs and efficacy of the decision criteria. Longer memory and higher learning rates have been found to result in a cheaper total cost. The number of fuzzy learning orders and the overall cost shift in an inversely proportionate manner as learning increases.The fuzzy learning case's total cost decreases rather than the fuzzy case's if the differential fractional order rises within the range. The crisp model, which derives minimal total cost, is relevant as a mathematical bound for the other two realistic models that incorporate demand volatility and are the main goals of this simulation. These models ameliorate the burden of implausibly perfect demand knowledge. This simulation is used to analyze the increases in the inventory system's output for each model. With learning, the inventory and total cost are less than those of a pure fuzzy system: the fuzzy model's delta over the crisp limiter is reclaimed by 62% for the order size and by 39% for the cost over the considered horizon. This accurately simulates how experience leads to better decision-making. A move to shorter memory has twin manifestations in order size: differential memory index generates linear reduction with a mean rolling fall of 10.5% per 0.1 change but the integral index causes a stronger (20%) quadratic reduction. For the total cost, their roles are reversed: both cause quadratic increase with shorter memory but the differential index has the stronger impact - a mean rolling rise of 11% per 0.1 change compared to 6.5% for the integral index. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
