Construction And Optimization Of Logistics Network For Sorted-waste Collection And Transportation System

Waste sorting is an issue that concerns people’s livelihood and sustainable social development.In recent years,source sorting of waste in China has become a new fashion and has received the attention and concern of the whole society.The traditional waste collection and transportation system with mixed transportation methods,waste drop-off locations,collection vehicles,and collection routes gradually failed to adapt to the pace of rapid reform.In its replacement,the waste sorting and collection system require that after the waste is sorted at the source,a new logistics network structure geared to the real needs of sorted waste drop-off,collection,and transfer is built.The transition from a traditional waste collection and transportation system to a waste sorting and transportation system is the focus and difficulty of waste sorting management.It is also a complex system engineering,which involves a series of NP-hard problems based on facility location,vehicle routing,location routing,etc.Its huge solution space makes the problem difficult to find the optimal solution in polynomial time.The design of optimization algorithms in complex scenarios of waste sorting presents new challenges in terms of solution quality and efficiency.Therefore,the scientific and rational establishment and optimization of the logistics network of waste sorting and transportation system is conducive to implementing the policy of waste sorting,improving its operational efficiency,and providing decision basis and technical support to promote the reform of waste sorting in China.In this thesis,the logistics network of waste separation and transportation system is explored to explore the reasonable planning and layout of sorting stations and the effective organization and arrangement of vehicle dispatching.Taking the actual demand as the starting point,multiple types of logistics network models are built according to the design ideas of independent optimization,simultaneous optimization,joint optimization,and two-echelon collaborative optimization of location and routing.The algorithm design,intelligent optimization,hyper-heuristic architecture,reinforcement learning,and other theories are used to design sequential,hierarchical,and integrated algorithms for the proposed models,which are dedicated to efficiently obtaining high-quality solutions to assist waste sorting management.Through experimental comparison and analysis of the characteristics and applicability of the different models,the management strategies and policy recommendations for waste sorting and transportation in China are summarized.The main research contents of this thesis are as follows.(1)Research on the current situation of urban waste sorting and collection in China.The collection and transportation mode,types of collection vehicles,future collection trends,and shortcomings in the collection and transportation process of municipal domestic waste sorting are explored and analyzed.Under China’s waste sorting policy,residents put out their waste at regular and fixed points,which requires the re-siting and design of community waste collection points and transfer stations.It puts forward higher requirements on the fairness and maximum coverage of the facilities.The need to independently design collection and transportation routes for different types of waste,with new elements of waste collection and logistics networks such as mixed periodicity,heterogeneous models,flexible multiple vehicles,queuing time,and collaborative scheduling all increase the difficulty of location and route planning.Based on this current research,we aim to establish the corresponding logistics network models for different stages of waste sorting and collection systems.We also explore the models and algorithms for siting multiple types of waste collection stations and planning the routes for waste sorting and collection under the sorting system.(2)Construction of logistics network model for waste sorting and collection system.Based on the current situation and research gaps in waste sorting and collection systems,the location and routing problems are explored in depth.The independent optimization,simultaneous optimization,joint optimization,and two-echelon collaborative optimization models are developed for the location and routing of sorted waste.In the independent optimization,two types of problems,namely,a model for waste transfer station location with capacity set coverage and a model for heterogeneous periodic mixed vehicle route with time windows are established.A green location-routing model with flexible multi-compartment in the simultaneous optimization,considering the factors of multi-compartment vehicles and carbon emission.In the joint optimization,the correlation between multiple location-routing problems is pointed out and a capacitated multiple location-routing model with queuing time is developed.In the two-echelon collaborative optimization,the synergistic relationship between waste treatment facilities is built,and the P-center location-routing two-echelon collaborative model is established based on location fairness and dispatch economy.(3)Design of logistics network optimization algorithm for waste sorting and collection system.For the various logistics network models established,efficient solution algorithms are constructed and good solutions are obtained according to the scale and characteristics of the problems.In the involved location problems,a reduced-order backtracking algorithm that can quickly reduce the size of the problem while obtaining an exact solution is proposed by studying and proving the mathematical properties of the problem for solution space pruning.The complex routing problem is decomposed into a multi-layer vehicle routing problem,and a new ant colony algorithm is developed that can exchange information and cooperate among groups.Under the hyper-heuristic algorithm architecture,Q-learning and multi-strategy-based,cross-entropy and simulated annealing-based hyper-heuristic algorithms are designed to solve the simultaneous optimization and joint optimization of the location-routing problem,respectively.A hybrid method for solving the two-echelon collaborative location routing is constructed by combining reduced-order backtracking,dynamic programming,and adaptive large neighborhoods.The performance of the designed algorithm is analyzed in terms of both optimization accuracy and computational efficiency.Their effectiveness is fully tested by comparing them with existing intelligent algorithms with the help of nonparametric statistical methods such as the Wilcoxon test.(4)Application of logistics network for waste sorting and collection system.According to the operation process of China’s waste collection and transportation system,several designed logistics networks are applied to the waste sorting and transportation system in Shanghai.A heterogeneous periodic hybrid vehicle routing model with time windows is applied in the sorting drop-off/collection phase.A simultaneous locationrouting optimization model is used in the waste transportation phase.A joint multilocation-routing optimization model is developed for the waste transfer stage.The synergy of transfer stations is introduced into the waste sorting and collection system.The two-echelon collaborative optimization model is used to unify and coordinate the logistics network of Shanghai’s waste sorting and transportation system,providing a scientific location and reasonable scheduling for it.The advantages,disadvantages,and applicability of several types of logistics network planning are explored using the ablation research method.The costs and operational efficiency of waste collection and transportation systems are analyzed.Policy recommendations and management insights for China’s urban waste sorting and collection systems are summarized by combining sensitivity analysis and experimental results to provide theoretical support for scientific and efficient waste sorting and collection work.To sum up,this thesis provides a variety of models and algorithms for the design and analysis of logistics networks of waste sorting and transportation systems.And an indepth study of the collaboration between different participants in the system is conducted.It can not only provide new means and options for solving related problems in waste management systems,improve their operational efficiency,and have certain theoretical value and practical significance for the study of waste sorting and transportation systems,but also develop characteristic and intelligent methodological tools for a series of interdisciplinary fields such as optimization theory and systems engineering in operations research and management,and provide a basis for decision-making for relevant departments.