Multi-temporal time window based path optimization for smart recycling bin removal vehicles

In the context of the digital economy, this study addresses the challenges associated with the scheduling of smart recycling bin emptying vehicles, particularly concerning arbitrary scheduling and high operational costs. A path optimization method is proposed that takes into account multi-time domain constraints within a single recycling cycle. This design leverages the real-time sensing capabilities of Internet of Things (IoT) enabled smart recycling bins. This paper introduces a combination of greedy algorithms, wolf pack algorithms, and variable neighborhood search algorithms to enhance the adaptive crossover, parallel crossover, and mutation processes, a hybrid genetic algorithm, termed the Multi-Temporal Time Window Algorithm (MGA-LTW), is developed. When evaluated against other IGA algorithms on a standardized platform and across various Solomon benchmark scenarios, the MGA-LTW algorithm exhibits a $2.5 \%$ reduction in the average objective value and a $16 \%$ decrease in average computation time. Moreover, the superior performance of the proposed approach is validated through the ablation experiments.In conclusion, the aforementioned method was implemented for optimizing the path of a real intelligent recycling bin reverse logistics clearance vehicle in Jinan. This application resulted in a cost reduction of $26.9 \%$ and a $45.5 \%$ decrease in vehicle utilization. These outcomes highlight the significant practical implications of the method and provide valuable decision-making support for enterprises aiming to enhance management efficiency and reduce operational costs.