Bi-objective perishable product delivery routing problem with stochastic demand. (January 2023)

• Record Type: Journal Article
• Authors: Wang, Qi; Li, Hui; Wang, Dujuan; Cheng, T.C.E.; Yin, Yunqiang
• Is Part Of: Computers & industrial engineering. Volume 175(2023)
• Publication Date: 2023-01
• Subjects: Perishable product -- Stochastic demands -- NSGA-II algorithm -- Adaptive large neighborhood search -- Routing; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2022.108837
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: A perishable product delivery routing problem with stochastic demand is studied. The goal is to minimize the expected operating cost and maximize the expected customer satisfaction. Two types of vehicles cooperate to serve the customers. A multi-object evolutionary algorithm combing NSGA-II with ALNS is developed. Experiment results show the developed algorithm performs well in finding the Pareto front. Abstract: In recent years, with the fast expansion of the market for perishable products, the transport of perishable products has become a critical issue in logistics research. Freshness and timeliness are two main factors affecting customer satisfaction with perishable product delivery, which are also closely related to the operating cost. To efficiently balance the operating cost and customer satisfaction, we consider the problem of bi-objective optimization with stochastic demand to find the Pareto-optimal solution of minimizing the total expected operating cost and maximizing the total expected customer satisfaction, where the demand is unknown until the delivery vehicle reaches the customer's location. Two types of vehicles cooperate to serve the customers, where one type is used to serve the initial requests of the customers, while the other type is used to fulfil the unmet demand caused by uncertain demand and non-compliance with the freshness requirement. We present a new evaluation method to measure customer satisfaction affected by time windows, which is more in line with customer psychology than the linear calculation method in existing studies. We develop a multi-objective evolutionary algorithm (MOALNS) that combines the NSGA-II algorithm with the adaptive large neighborhood search to solve the problem, where the expected operating cost of a given route under the vehicle cooperation strategy is minimized by a dynamic programming algorithm. Extensive numerical studies show that MOALNS performs well in finding the Pareto front and reveal the impacts of the key model parameters on the optimal solution, providing insights for the decision-maker to devise delivery schemes that trade-off the operating cost for customer satisfaction.