Long-term investment planning of linking energy and water technologies: a hybrid optimization and deep neural network approach

The aim of this project is to develop a data-driven, decision support framework to investigate short-run operational scheduling and long-term investment planning problems for decentralized (linking)water and energy technologies (such as water treatment plant, water desalination units) in the context of integrated electricity water and system. The decisions made are inferred by information extracted directly from real-world and expert knowledge. To that end, we will look into novel optimization based as well as statistical (machine) learning algorithms that will be tailored to capture physical phenomena and to describe the underlying uncertainties with an adequate accuracy. The uniqueness of the framework lies in innovative fundamental research performed and robustness of the use-case agnostic solution methods that will be developed. That is, novel solution methods are proposed considering the technical characteristics of approaches developed in WP6 of the Aqua-Connect project. In addition, we provide generic knowledge and techniques applicable to different case studies in the end or as follow up of the project.