Advancing simulation-optimization modeling (SOM) techniques for the development of robust air and water pollution load reduction-allocation (PLRA) programs

推进模拟优化建模 (SOM) 技术，以开发稳健的空气和水污染负荷减少分配 (PLRA) 计划

• 批准号: 251325-2011
• 负责人: Liu, Lei
• 金额: $ 1.38万
• 依托单位: Dalhousie University
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2014
• 资助国家: 加拿大
• 起止时间: 2014-01-01 至 2015-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=547664
• 关键词: Advancing; simulation; optimization; modeling; SOM

Deteriorations of regional ambient air quality and natural surface water bodies are among the major environmental concerns facing decision makers and stakeholders. Effective air and water pollution load reduction-allocation (PLRA) programs are desired for air/water quality restoration and management. Simulation-optimization modeling (SOM) techniques play a central role in developing such PLRA programs, where the simulation model serves as a numerical representation of the prototype nonlinear air/water system to predict the effects of external forces (i.e. pollutant loads) on the internal responses (i.e. air/water quality); and optimization models are coupled to iteratively drive the simulation model to find the optimal PLRA strategies. The primary objective of this proposed research aims to develop advanced SOM techniques for a PLRA simulation-optimization analysis; it entails: (a) characterize the nonlinearities and uncertainties of complex air and water systems, (b) propose nonlinearity-to-uncertainty mapping (N2UM) approaches to convert the original nonlinear SOM model to its uncertain LP equivalence and investigate its performance over existing functional approaximators, (c) develop advanced inexact optimization algorithms for solving LP-equivalent models under multiple uncertainties, and (d) develop interactive post-optimality approaches to reflect the tradeoffs between decision risk and system safety for generating operational decision schemes. The developed SOM techniques will be applied to real-world cases within a Canadian air and water context. The proposed approaches represent a unique contribution to methodologies in the PLRA field. It could provide technical supports for tackling challenges in terms of computation bottleneck, uncertainty handling, and decision-risk & system-safety tradeoff analysis. Its applications to Canadian cases will help Canadian stakeholder improve their effectiveness and efficiency in managing air/water pollution problems. Besides, a number of graduate students and PDFs will be trained through participating in the proposed research. The trained HQPs will fill a pressing need for environmental management professionals in various sectors.

区域环境空气质量和自然地表水的恶化是决策者和利益相关者面临的主要环境问题之一。空气/水质恢复和管理需要有效的空气和水污染负荷减少分配（PLRA）方案。仿真优化建模（SOM）技术在开发此类PLRA程序中发挥着核心作用，其中仿真模型作为原型非线性空气/水系统的数值表示，以预测外部力量（即污染物负荷）对内部响应（即空气/水质量）的影响；与优化模型相耦合，迭代驱动仿真模型，寻找最优PLRA策略。本研究的主要目标是为PLRA仿真优化分析开发先进的SOM技术；它需要:(a)表征复杂空气和水系统的非线性和不确定性；(b)提出非线性到不确定性映射（N2UM）方法，将原始非线性SOM模型转换为不确定LP等价，并研究其在现有函数逼近器上的性能；(c)开发先进的不精确优化算法，用于求解多重不确定性下的LP等效模型。(d)开发交互式后最优性方法，以反映决策风险和系统安全之间的权衡，以生成运营决策方案。开发的SOM技术将应用于加拿大空气和水环境中的实际案例。所提出的方法是对PLRA领域方法的独特贡献。它可以为解决计算瓶颈、不确定性处理、决策风险与系统安全权衡分析等方面的挑战提供技术支持。将其应用于加拿大的案例将有助于加拿大利益相关者提高管理空气/水污染问题的有效性和效率。此外，一些研究生和pdf将通过参与拟议的研究得到培训。训练有素的环境管理人员将填补各行各业对环境管理专业人员的迫切需求。