Abstract:

To address the challenges of inconvenient transportation and weak disaster prevention capabilities in closed areas with complex terrain， an optimization model of drone distribution point location and flight route planning is proposed for “dual use of usual and emergency situations”. The model is based on the emergency and emergency reconstruction of food security infrastructure in the closed area with complex terrain. Firstly， the investment and operation mode of drone distribution system， the selection of drone distribution mode and the selection of distribution point construction mode are discussed. Secondly， based on the selection of the above mode， the full life cycle cost of food infrastructure construction and operation is considered. The construction cost considered in this system includes the cost of transforming the distribution point to adapt to the operation of drone system， the cost of expansion due to insufficient capacity of the distribution point after the new demand is generated， and the cost of drone system due to the addition of new drone and other accessories. The operation cost considered includes the loss cost of drone system， the labor cost due to the addition of porters， the power cost and water cost due to the daily operation of the distribution point and drone transportation； In case of emergency， the main objective of the optimization model constructed in this paper is to minimize the distribution time， meeting the efficiency demand of emergency material distribution， without considering other costs. The main optimization objective is to minimize the sum of construction cost and operation cost of the drone “dual use of usual and emergency situations” distribution system， and reduce the distribution time in case of emergency. To achieve the optimization goal， an improved chaotic adaptive genetic algorithm （ICAGA） is proposed in this paper. Firstly， the joint optimization framework， solution model and optimization algorithm of the “dual use of usual and emergency situations” drone food distribution system are combined by the “dual use of usual and emergency situations” integrated fitness solution method. Then the PWLCM chaotic mapping and the introduction of multi module improved adaptive genetic algorithm are used to improve the optimization speed of genetic algorithm and the ability to jump out of the local optimal solution. Finally， the effectiveness and reliability of the proposed drone emergency and emergency distribution model are verified by a real case in BZ Town， which achieves the goal of “dual use of usual and emergency situations” and maximizes economic benefits at the same time. At the same time， by comparing with the traditional genetic algorithm， chaotic genetic algorithm and adaptive genetic algorithm， it is found that the ICAGA algorithm proposed in this paper performs better in solving speed and the final optimal solution. In addition， an additional disaster simulation test is conducted to verify the ability of drone distribution system based on the model to ensure material supply in the event of disasters. Through the real case verification， disaster simulation test and sensitivity analysis of BZ Town， it is found that the optimization model proposed in this paper can maximize the economic benefits of the “dual use of usual and emergency situations” drone distribution system， and has the robustness to respond to disasters and demand changes， which provides strong support for the construction of a more comprehensive drone distribution system.

Key words: dual use of usual and emergency situations, drone delivery, improved chaos adaptive genetic algorithm, joint optimization