Development of Advanced Tools for Multi-Objective Analysis of Water Resources Engineering Systems

水资源系统本质上是多用途的，任何合理的环境管理决策，如如何操作它们以减轻未来气候条件的影响，都应反映它们的多目标性质。例如，河流-水库系统的运行可以设计为最大限度地提高城市、农业和工业取水的可靠性，最大限度地从水力发电和商业航运中获得货币效益，并最大限度地减少对系统中栖息地的负面环境影响。我的研究项目产生了先进的工具来分析水资源工程系统在一个高效和有效的多目标模拟优化框架。这些工具将利用高性能计算资源来适应广泛的计算规模，并将使研究人员、高级建模者和环境系统分析师能够访问。在我的研究计划中开发的工具将适用于广泛的水资源工程问题，包括河流-水库运行，配水网络设计和水文模型校准。这项研究计划将特别有利于加拿大，加拿大在研究中利用了主要的水资源案例研究。我的研究计划的一个例子案例研究将是五大湖-圣劳伦斯河-水库系统，它对加拿大和美国的经济非常重要，并且具有水库运行，水文建模和模型校准组件。另一个例子是由马尼托巴水电公司使用近似技术将系统行为线性化的高度复杂和非线性的水力发电系统。在我的研究计划中开发的工具不需要线性化问题。因此，他们将能够在优化非线性模型中精确地模拟系统行为，并找到更好的解决方案，以最大限度地提高运营效益（目前超过20亿美元/年），最大限度地降低维护成本，最大限度地提高产能扩张效益和/或最大限度地减少运营对环境的影响。在这个为期5年的研究计划中，学生将获得宝贵的经验，开发和实施创新的算法，以解决市、省、国家和国际层面的水资源工程问题，并成为加拿大未来水资源系统分析领域的一些专家。

Water resources systems are inherently multipurpose, and any sound environmental management decision such as how to operate them to mitigate impacts of future climate conditions should reflect their multi-objective nature. For example, the operation of a river-reservoir system can be designed to maximize the reliability of supplying water for urban, agricultural, and industrial intakes, to maximize the monetary benefits obtained from hydropower generation and commercial navigation, and to minimize negative environmental impacts on habitats in the system. My research program generates advanced tools to analyze water resources engineering systems in an efficient and effective multi-objective simulation-optimization framework. These tools will utilize high performance computing resources to be adaptable to a wide range of computational scales and will be made accessible to researchers, advanced modellers, and environmental system analysts. Tools developed in my research program will be applicable to a wide range of water resources engineering problems including river-reservoir operation, water distribution network design and hydrologic model calibration. This research program will specifically benefit Canada that has major water resources case studies utilized in research. An example case study of my research program will be the Great Lakes - St. Lawrence river-reservoir system that is very important to the Canadian and US economies and has reservoir operation, hydrologic modeling, and model calibration components. Another example will be the highly complex and non-linear hydropower generating system operated and maintained by Manitoba Hydro using approximation techniques that linearize the system behavior. Tools developed in my research program do not need to linearize the problem. As such, they will be able to incorporate in the optimization non-linear models that accurately simulate the system behavior and find better solutions that maximize the operation benefits (currently more than $2 billion/yr), minimize maintenance costs, maximize capacity expansion benefits and/or minimize environmental impacts of the operation. Students who will be trained over the course of this 5-year research plan will gain valuable experience to develop and implement innovative algorithms for solving water resources engineering problems at the municipal, provincial, national and international levels and become some of Canada's future experts in the field of water resources systems analysis.