Water digitisation is a global trend in the water industry. Although hundreds/thousands of sensors have been deployed in different water systems (e.g., water distribution systems, sewer networks, stormwater systems), it is not enough to monitor the whole system given its sheer size. Numerical models were used to gain more insights, but they may involve large errors due to model uncertainties. Overall, pure physical monitoring or numerical modelling is unable to enable a full-scale monitoring. Artificial Intelligence has been rapidly applied in water networks. For example, the physics-informed neural network was recently applied in water systems. This project will develop novel AI approaches to integrate data (from sensors) with physics (from physics-based models) to enable full-scale state estimation in urban water systems. The use of physics ensures an explainable AI guided by classical hydraulics. The use of data incorporates hidden information into the AI to eliminate parameter uncertainties.
- Student type: International, Domestic
- Research degree type: PhD
- Signature research theme: Sustainable Green Transition
- Supervisor: Dr Wei Zeng
