Engineering Water Pollution Control System Design Based on Robust Optimization Strategy

2018 International Conference on Advanced Electronic Materials, Computers and Materials Engineering

材料科学;无线电电子学;计算机科学

Zhang, Zhuorong^1

Department of Civil Engineering, Auburn University, Auburna

AL

36849, United States^1

关键词: Application capability;    Evaluation and selections;    Latin hypercube sampling;    Multi-attribute evaluations;    Multiple-objectives;    Robust optimization models;    Robustness analysis;    System optimizations;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/439/5/052007/pdf
DOI  :  10.1088/1757-899X/439/5/052007

来源: IOP

【 摘 要 】

Environmental and health problems caused by engineering water pollution have become an important factor restricting the sustainable development of our country's engineering. Our country's engineering water pollution control has the characteristics of diverse structure, multiple objectives, related elements and great uncertainty, which brings challenges to the reasonable and effective development of engineering water pollution control planning. The structural design and technical selection method of engineering water pollution control established in this paper includes two parts: multi-objective robust optimization model and robust analysis framework. Among them, the multi-objective robust optimization model uses Latin hypercube sampling to improve the non-dominated sequencing genetic algorithm to provide an optimal scheme set that takes robustness into consideration for the planning of the engineering water pollution control system. A multi-attribute evaluation index system is constructed in the framework of robustness analysis. Through evaluation and selection of the optimal scheme set, subjectivity of system optimization scheme selection is reduced, and the scientific and quantitative level of scheme evaluation is improved. Based on the uncertainty analysis, key factors and sensitive areas affecting the robustness of the system are identified, which further improves the interpretation and application capabilities of the multi-objective robust optimization model.