Robust Optimization For Structure Design And Technology Selection Of Urban Water Pollution Control System

The environmental and health problems incurred by urban water pollution have become a big crisis confronting China’s sustainable urban development.Urban water pollution control system in China is characterized bycomplicated system structure, multiple objectives, many correlative elements and lots of uncertainties. These features have posed tremendous challenges for effective planning of urban water pollution control. To solve these problems caused by the need of integrated optimization of system structure and technology combination, point source and nonpoint source pollution, nominal and robust performances, an integrated method for system structure and technology selection ofurban water pollution control is developed based on robust optimization to support system planning.A technology database is constructed and stores efficiency information of unit treatment process and technology synthesis rules as well. Atreatment train assembling module is nested in the database, which can generate alternative treatment trains for later technology combination selection by invoking technology performance parameters and synthesis rules. The method of system structure design and technology selection includes multi-objective robust optimization model and robustness analytical framework. The multi-objective robust optimization model, which is developed based on the Latin hypercube sampling(LHS) improved non-dominated sorting genetic algorithm(NSGA_II), can generate Pareto optimal solutions set which considers the environment, economic and robustness performances. The robustness analytical framework including a multi-attribute evaluation index system to evaluate comprehensive performances of optimal solution set, which improves the rationality and the quantitative level, and further reduces the subjectivity in decision making. Afterwards uncertainty analysis is applied in the framework to identify critical factors and sensitive areas that effluent robustness performance.The established method and developed tools areimplemented to a hypothetical representativecity to provide multi-objective, multi-layer optimization solutions under different constraints and uncertainties. Comprehensive evaluation is carried out to providesupportive information for decision making, and further identify the critical paramters, as well as the complex relationship of system structure, treatment combination and system performances. This research can provide valuable quantitative assessments to support the planningof urban water pollution controland the development trend of treatment technologies. The complex relationship between environmental, economic and robust performances can be revealed by our study, as well as the features of uncertainties which transfer and accumulation within the model.