考慮多房型飯店之批量訂單預訂控制

Abstract

本篇論文在收益管理領域中研究飯店預訂控制策略，討論時效性商品的產能分配問題。飯店的獲利空間受限於固定的產能，然考量消費族群特性之不同，管理者將消費者分為較早下訂單且享有折扣的旅行社以及較晚訂房不具優惠的散客，透過差別訂價最大化房間的實現價值以提升利潤。本研究討論飯店管理者面對需求變動且具不確定性，在給定一批量的旅行社訂單中，將決定接受的訂單子集以及升級策略，以減緩供需不平衡。我們納入旅行社訂單取消的隨機性，並採用超額預訂策略，透過制定散客預訂額度，控管超量需求。我們建構混整數模型以描述四種環境設定的飯店預訂控制問題，由於問題所需的運算資源量大，我們提出演算法並使用 Gurobi 為比較基準，生成不同情境的資料集，評估演算法的表現與穩定性，演算法在合理的運算時間內能提出足夠好的解。

We study hotel booking control strategies for capacity allocation in revenue management to maximize the availability of perishable products. Hotels generally sell rooms to different customer groups. The profits are tied up to the fixed capacity and hinge on how to sell them optimally. While travel agents place bulk orders at favorable prices with a longer lead time, individual travelers book rooms at high marginal revenue. In this study, an accommodation provider with a batch of agent orders on hand composes a subset of orders to accept and propose upgrade arrangements to tune their requirements. With delicate operations to maximize the expected profits, the manager would alleviate mismatches between supply and sum of stochastic individual booking demands and aggregate capacity consumption by agents. In the presence of uncertain cancellations, we implement overbooking strategies and set the maximum accessible booking amount for individuals to trim excess demands. We formulate mixed-integer programs to describe the problems and propose a heuristic algorithm. We employ Gurobi Optimizer as a benchmark and simulate datasets with different scenarios to evaluate the performance and robustness of our algorithm. Our algorithm guarantees good enough solutions by a reasonable time in our experiments.