基于改进多目标蜣螂优化算法的生鲜多温共配路径优化研究

Research on Optimization of Fresh Produce Multi-Temperature Joint Distribution Path Based on Improved Multi-Objective Dung Beetle Optimization Algorithm

摘要: 考虑到生鲜配送场景下消费者需求零散化、个性化和及时性的特点，本文开展了生鲜多温共配路径问题研究。首先，基于客户对生鲜配送时间是否及时准确的敏感程度，引入时间抵制度模型，构建以运输成本最小、客户时间抵制度最小为目标的生鲜多温共配路径多目标优化模型。其次，针对蜣螂优化算法的改进，设计了一种以tent混沌映射与逆向学习策略生成初始种群，在蜣螂觅食阶段引入自适应步长策略与凸透镜成像策略，滚球阶段引入曲线自适应黄金正弦策略的多策略改进的多目标蜣螂优化算法用于模型求解。最后，通过算例分析验证模型和算法的有效性，与未改进的多目标蜣螂优化算法进行对比，验证了改进多目标蜣螂优化算法性能的优越性。

Abstract: Considering the characteristics of decentralized, personalized, and timely consumer demand in the context of fresh produce delivery, this paper investigates the multi-temperature joint distribution path problem for fresh produce. Firstly, based on customers’ sensitivity to the timeliness and accuracy of fresh produce delivery, a time resistance model is introduced. A multi-objective optimization model for the multi-temperature joint distribution path of fresh produce is constructed with the objectives of minimizing transportation costs and customer time resistance. Secondly, for the improvement of the dung beetle optimization algorithm, a multi-strategy improved multi-objective dung beetle optimization algorithm is designed. This algorithm generates an initial population using tent chaotic mapping and inverse learning strategy, introduces an adaptive step size strategy and a convex lens imaging strategy in the dung beetle foraging phase, and incorporates a curve adaptive golden sine strategy in the rolling phase for model solving. Finally, through case analysis, the effectiveness of the model and algorithm is verified. Compared with the unimproved multi-objective dung beetle optimization algorithm, the superiority of the improved multi-objective dung beetle optimization algorithm’s performance is validated.