Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Robust distributed MPC design for energy hub network with data-driven demand forecasting
In this project, we aim to design a forecasting model to predict the demand of an energy hub using data driven techniques and formulate a robust MPC for the distributed optimization of a network of energy hubs.
Keywords: Optimization, Large-scale systems, Energy hub, Data driven modelling, energy demand forecasting,
machine learning, model predictive control, multi-energy systems
Energy hubs are multi-carrier energy systems that incorporate conversion, storage and network technologies of different energy carriers such as electrical, thermal and/or chemical processes that follow individual demand and supply patterns under distinct time scales. Energy hubs facilitate easier integration of intermittent renewable energy sources, greater flexibility and higher efficiency of the system. Multiple energy hubs in an interconnected network with energy trading between them further amplifies the systemic and economic benefits of individual energy hub by sharing information on their capabilities, intended production and consumption, such that they can be matched locally.
The cooperative operation of the network using model predictive control strategies rely on future demand profiles for the energy and heating demand of the hubs. Data Driven modelling approaches are increasingly utilized since first principles models may be costly to develop, difficult to maintain and in some cases, first principle identification methods cannot be effectively employed. The prediction model that relies on data for its forecast can be integrated into the MPC optimization framework for online correction in the dispatch problem based on the past prediction error. A plug and play framework that models a network of energy hubs has been designed along with a novel distributed MPC algorithm for the multi-objective optimization of the network. Integrating uncertainty of the forecast into the optimization scheme using robust or stochastic optimization techniques improves the results of the MPC scheme and the overall operation of the energy hub network.
The goal of this project is two fold. Firstly, design and evaluate different data driven techniques to model demand patterns of an energy hub. The developed method should account for the dependence of a load on external factors such as temperature, price, etc. Suitable methods should be identified, implemented and where necessary, adapted. The second part of this project is to incorporate the demand forecasting model into an MPC framework. The distributed model predictive controller utilizes the forecasted demand to optimize the operation for the network. Robust and stochastic optimization methods should be used in the MPC to account for the uncertainity and variance in the forecasted demand. Finally, to design an online correction algorithm that corrects the prediction at every MPC time step based on the error in the forecast at the last elapsed time step. The forecasting model and the control algorithm should be tested in simulation, as well as on real data of the NEST demonstrator at Empa.
Energy hubs are multi-carrier energy systems that incorporate conversion, storage and network technologies of different energy carriers such as electrical, thermal and/or chemical processes that follow individual demand and supply patterns under distinct time scales. Energy hubs facilitate easier integration of intermittent renewable energy sources, greater flexibility and higher efficiency of the system. Multiple energy hubs in an interconnected network with energy trading between them further amplifies the systemic and economic benefits of individual energy hub by sharing information on their capabilities, intended production and consumption, such that they can be matched locally.
The cooperative operation of the network using model predictive control strategies rely on future demand profiles for the energy and heating demand of the hubs. Data Driven modelling approaches are increasingly utilized since first principles models may be costly to develop, difficult to maintain and in some cases, first principle identification methods cannot be effectively employed. The prediction model that relies on data for its forecast can be integrated into the MPC optimization framework for online correction in the dispatch problem based on the past prediction error. A plug and play framework that models a network of energy hubs has been designed along with a novel distributed MPC algorithm for the multi-objective optimization of the network. Integrating uncertainty of the forecast into the optimization scheme using robust or stochastic optimization techniques improves the results of the MPC scheme and the overall operation of the energy hub network.
The goal of this project is two fold. Firstly, design and evaluate different data driven techniques to model demand patterns of an energy hub. The developed method should account for the dependence of a load on external factors such as temperature, price, etc. Suitable methods should be identified, implemented and where necessary, adapted. The second part of this project is to incorporate the demand forecasting model into an MPC framework. The distributed model predictive controller utilizes the forecasted demand to optimize the operation for the network. Robust and stochastic optimization methods should be used in the MPC to account for the uncertainity and variance in the forecasted demand. Finally, to design an online correction algorithm that corrects the prediction at every MPC time step based on the error in the forecast at the last elapsed time step. The forecasting model and the control algorithm should be tested in simulation, as well as on real data of the NEST demonstrator at Empa.
1. Literature review: study the existing data driven and grey-box modelling approaches for multienergy systems and building systems to understand the different methods for demand forecasting. To understand the concepts of robust/stochastic MPC and the application of MPC in integrated energy system.
2. Investigation and comparison of different forecasting methods and cross-validating on historical data. Consideration of seasonal and inter dependency of the forecast and prediction of different types of loads as well as an estimate of the variance and uncertainity in the forecast.
3. Integrate the demand prediction into the MPC framework and design an online correction algorithm that corrects the output based on the error in the forecast in the previous time step.
4. Design of the robust MPC algorithm to integrate uncertainty of the forecast into the optimization scheme improves the results of the MPC scheme.
5. Experimental validation of the data driven demand forecasting and the control scheme with online error correction algorithm on a real life energy hub system.
6. Documentation and final presentation of the work
1. Literature review: study the existing data driven and grey-box modelling approaches for multienergy systems and building systems to understand the different methods for demand forecasting. To understand the concepts of robust/stochastic MPC and the application of MPC in integrated energy system. 2. Investigation and comparison of different forecasting methods and cross-validating on historical data. Consideration of seasonal and inter dependency of the forecast and prediction of different types of loads as well as an estimate of the variance and uncertainity in the forecast. 3. Integrate the demand prediction into the MPC framework and design an online correction algorithm that corrects the output based on the error in the forecast in the previous time step. 4. Design of the robust MPC algorithm to integrate uncertainty of the forecast into the optimization scheme improves the results of the MPC scheme. 5. Experimental validation of the data driven demand forecasting and the control scheme with online error correction algorithm on a real life energy hub system. 6. Documentation and final presentation of the work
Varsha N. Behrunani varsha.behrunani@empa.ch or bvarsha@ethz.ch
Varsha N. Behrunani varsha.behrunani@empa.ch or bvarsha@ethz.ch