Probability-based adaptive capacity rental strategy on shared platform with unknown demand distribution

Capacity sharing presents a transformative strategy in manufacturing, driven by the increasing demand for flexibility and efficiency in a highly uncertain market. Shared platforms play a crucial role in facilitating this transformation by offering a variety of scenarios that enable enterprises to make adaptable decisions. This paper develops a capacity sharing model on a shared platform, addressing two scenarios-standardized and differentiated scenarios-the latter incorporating cost discounts. We propose a probability-based adaptive rental strategy (PAS) in the absence of demand distributions. This strategy depicts human psychology and behavior through three steps: designing options, calculating probabilities, and establishing schemes. It differs from direct optimization of decisions by adaptively addressing stochastic problems through options and probabilities. Experiments demonstrate that PAS can balance flexibility and stability across diverse environments, including Poisson, Normal, multimodal, heavy-tailed distributions, and real-world datasets. Furthermore, it achieves near-optimal average profit performance, with improvements attainable through option adjustments.