Revenue optimisation approach for auction logistics centre: an investigation of Chinese flower market

This research is motivated by China's second largest flower auction market. Its existing auction logistics plan is to group the auction trolleys in a fixed order of 'same category and same grade'. The auction logistics centre cannot respond to demand fluctuations by dynamically optimising the order of participation, resulting in overall trading revenue decline. This paper proposes an integrated decision-making framework based on the theory of revenue optimisation. Through effective demand forecasting, the demand curves of different types of auction lots in multi-period are obtained. In the face of the reserved order of a limited auction trolley, order acceptance strategy is adopted to flexibly allocate auction trolleys, with the goal of revenue maximisation. Considering the impact of demand uncertainty and forecast errors on system performance, a robust control is employed. Through designing a genetic-variable neighbourhood algorithm, the solving of this large-scale practical problem can be controlled within acceptable computation time. Numerical experiments verify that the proposed approaches are superior to the traditional first-come-first-served rule. Sensitivity analyses provide significant implications for the scientific decision-making of managers, including the allocation ratio of group orders to individual orders, discounts on entry fees for group orders, and the division of auction time periods.