An Optimization-Based State Estimation Framework for Large-Scale Natural Gas Networks
[1];
[2]
- Department of Chemical and Biological Engineering, University of Wisconsin−Madison, 1415 Engineering Drive, Madison, Wisconsin 53706, United States; Global Security Sciences Division, Argonne National Laboratory, 9700 South Cass Avenue, Lemont, Illinois 60439, United States
- Department of Chemical and Biological Engineering, University of Wisconsin−Madison, 1415 Engineering Drive, Madison, Wisconsin 53706, United States
We propose an optimization-based state estimation framework to track internal spacetime flow and pressure profiles of natural gas networks during dynamic transients. We find that the estimation problem is ill-posed (because of the infinite-dimensional nature of the states) and that this leads to instability of the estimator when short estimation horizons are used. To circumvent this issue, we propose moving horizon strategies that incorporate prior information. In particular, we propose a strategy that initializes the prior using steady-state information and compare its performance against a strategy that does not initialize the prior. We find that both strategies are capable of tracking the state profiles but we also find that superior performance is obtained with steady-state prior initialization. We also find that, under the proposed framework, pressure sensor information at junctions is sufficient to track the state profiles. We also derive approximate transport models and show that some of these can be used to achieve significant computational speed-ups without sacrificing estimation performance. We show that the estimator can be easily implemented in the graph-based modeling framework Plasmo.jl and use a multipipeline network study to demonstrate the developments.
- Research Organization:
- Argonne National Lab. (ANL), Argonne, IL (United States)
- Sponsoring Organization:
- USDOE Office of Electricity (OE)
- DOE Contract Number:
- AC02-06CH11357
- OSTI ID:
- 1439882
- Journal Information:
- Industrial and Engineering Chemistry Research, Vol. 57, Issue 17; ISSN 0888-5885
- Publisher:
- American Chemical Society (ACS)
- Country of Publication:
- United States
- Language:
- English
Similar Records
Enhanced data efficiency using deep neural networks and Gaussian processes for aerodynamic design optimization
Enhanced Data Efficiency Using Deep Neural Networks and Gaussian Processes for Aerodynamic Design Optimization