Functional Division Time Window-based Vehicle Routing Optimization Strategies For Collecting And Transporting MSW Streams In High-Tech Industrial Development Zone

The tradeoff among environmental, economic, and social impact in solid waste management system (SWMS) brings about challenges for making such effective strategies, especially for waste collection and transportation. Thus, it is indispensable for decision makers to consider different options using an optimization method. However, due to availability and quantifiability of related information, extensive uncertainties exist in many system components and impact factors, leading to increased complexities in the related planning efforts and affecting consequent decision processes. That the questions of how the SWMS should be managed from a sustainable perspective, the proper formulation of optimization methods or system assessment tools with multifaceted features connecting all factors together are becoming indispensable in the future work.One spatial characteristic of high-tech industrial development zone is the function zoning, and based on that, the primary transport stage in municipal solid waste (MSW) Collection and Transportation Systems associated with this feature considered in the paper. The mathematical model in this thesis takes full account of the relationship between the sensitive time period, the function zoning and the vehicle routing by introducing the concept of time-window for the transportation node. Under these constraints, the optimization of the primary transport stage can be modeled as a VRP with time-window, intermediate facilities, and multiple disposal trips. Then, this paper expounds the designs of two heuristic constructive methods (extended insertion algorithm-ES and genetic algorithm-GA) based on function zoning, and analyses the defects of this two algorithms from two respects. On the basis of it, a multi-objective genetic algorithm (MOP+GA) is formulated which giving priority attention to negative social effects and the economic costs can be considered as well. In addition, the degree of polymerization of collection point in one vehicle route (Er) and the spatial independence of each vehicle route (Nr) are designed to measure the negative social externality in waste transportation. Furthermore, the Pareto ranking scheme and the Graham scanning technique are introduced for designing the new algorithm. And in this way, both the independence for each optimization goal and the good search characteristic for the optimal solution would be secure, and each vehicle route forms a convex hull by this new procedure. Thus, the multi-objective genetic algorithm in this thesis builds and takes into account a hierarchy of the different optimization goals by further exploring the high correlation between the primary transport stage and the function zoning in high-tech industrial development zone.In this case study, three designed heuristic algorithms are used to solve the optimization strategies problem for collecting and transporting MSW streams in such instance, and the visual simulation of the optimal path and convex hull is realized at last. Through the comparison and analysis of the three optimization schemes, the multi-objective genetic algorithm is superior to the other two formulations in fully consideration of functional division characteristics in reducing the negative social effects caused by high-frequency collection and the overlap of the vehicle route, especially under the time window constraint. Comparison results indicate that, in term of Nr, the new method is 20% and 29.41% higher than that of the extended insertion algorithm and genetic algorithm respectively. And in term of Er, It's 3.41% higher than that of the genetic algorithm. Furthermore, the new method is increased by only 0.97% higher than that of the genetic algorithm in term of unit cost, and saves 7.0?/t and 18.87% compared with the current scheme. In comparison with the negative social effects, this paper explained the cause of the low Er in the scheme of the extended insertion algorithm by combining the number of vehicles and convex hulls, and this interpretation also further evidence the limitation of the extended insertion algorithm. Finally, the sensitivity of the parameter analysis indicate that the economic cost in the scheme of the extended insertion algorithm and genetic algorithm has close correlation with the function zoning time-window proposed in this paper, While the multi-objective genetic algorithm perform well in reducing the negative social effects and economic cost with different setting of parameters, which prove the applicability of the new algorithm.This thesis extends the analysis for spacial feature in waste management, and represents a potential breakthrough for the existing MSW collection and transportation system. The proposed optimization strategies which are different from than that in regular cities will promote a more rational approach for optimizing collecting and transporting MSW streams and reducing the NIMBY movement in high-tech industrial development zone and similar areas.