Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Model Identification of Inverter-Interfaced Equipment in Transmission and Distribution Systems
This project is on developing methods for modeling and identifications of inverter-based power resources. Moreover, we design suitable specific experiments for this identification procedure. Finally, we derive model validation techniques for assessing the derived models.
Keywords: Energy, Inverter-based Resources, Modeling and System Identification
The energy transition towards a higher presence of inverter-based resources (IBR) in power grids gives rise to a number of challenges of power system operation and planning. One of them is understanding the dynamic behavior and analyzing system stability. For traditional power systems with rotating machines as energy conversion devices, the mechanisms of various instability phenomena are well understood and characterized. On the other hand, power electronics-based converters possess dynamic characteristics different from those of rotating machines. They are greatly affected by the underlying technology (e.g. wind turbines, PV panels, batteries, etc.), the presence of high order output filters, and various control loops, including nonlinear control. This results in dynamic phenomena not previously observed in power systems. To perform the analysis of how IBR could affect system stability and to better understand the behavior system operators and planners require high-fidelity IBR models. However, often such models are either unavailable or proprietary to the manufacturer,i.e., "black box". A common workaround is to use so-called generic models with a clearly defined structure. However, matching their response to that of a black-box model or an actual device can be a challenge, as it often requires expert tuning of model parameters.
The global objective of the project is the development of model identification methods for inverter-interfaced generation from measurements or simulated response. Of particular interest is the "grey-box identification" approach, where a certain structure of the identified model is assumed. This is motivated by the fact that many IBRs have common power elements regardless of the technology. For instance, grid-side DC-AC inverters are generally voltage source converters with an output filter with a typical topology of an L, LC, or LCL filter. Therefore, in many cases, an approximate IBR structure can be correctly determined or guessed from experience. The identification problem in that case becomes a problem of parameter estimation. Additionally, such an approach can be easily extended to defining a set of typical IBR structures. The goal of identification, in this case, is to determine a topology, along with the parameters, that is best fit to represent the dynamic response of the device. On the other hand, determining the control loop structure is less straightforward. Therefore, a more suitable strategy of black-box identification of the feedback transfer function matrix could be used. It remains to be seen how to extend these methods to nonlinear control.
Overall, the described "grey-box" approach is likely to be iterative and may require a set of heuristics along the identification process. Thus, it will not be as general as black-box identification methods but will result in a much more useful model.
For more details see the attached document.
The energy transition towards a higher presence of inverter-based resources (IBR) in power grids gives rise to a number of challenges of power system operation and planning. One of them is understanding the dynamic behavior and analyzing system stability. For traditional power systems with rotating machines as energy conversion devices, the mechanisms of various instability phenomena are well understood and characterized. On the other hand, power electronics-based converters possess dynamic characteristics different from those of rotating machines. They are greatly affected by the underlying technology (e.g. wind turbines, PV panels, batteries, etc.), the presence of high order output filters, and various control loops, including nonlinear control. This results in dynamic phenomena not previously observed in power systems. To perform the analysis of how IBR could affect system stability and to better understand the behavior system operators and planners require high-fidelity IBR models. However, often such models are either unavailable or proprietary to the manufacturer,i.e., "black box". A common workaround is to use so-called generic models with a clearly defined structure. However, matching their response to that of a black-box model or an actual device can be a challenge, as it often requires expert tuning of model parameters.
The global objective of the project is the development of model identification methods for inverter-interfaced generation from measurements or simulated response. Of particular interest is the "grey-box identification" approach, where a certain structure of the identified model is assumed. This is motivated by the fact that many IBRs have common power elements regardless of the technology. For instance, grid-side DC-AC inverters are generally voltage source converters with an output filter with a typical topology of an L, LC, or LCL filter. Therefore, in many cases, an approximate IBR structure can be correctly determined or guessed from experience. The identification problem in that case becomes a problem of parameter estimation. Additionally, such an approach can be easily extended to defining a set of typical IBR structures. The goal of identification, in this case, is to determine a topology, along with the parameters, that is best fit to represent the dynamic response of the device. On the other hand, determining the control loop structure is less straightforward. Therefore, a more suitable strategy of black-box identification of the feedback transfer function matrix could be used. It remains to be seen how to extend these methods to nonlinear control. Overall, the described "grey-box" approach is likely to be iterative and may require a set of heuristics along the identification process. Thus, it will not be as general as black-box identification methods but will result in a much more useful model.
For more details see the attached document.
The challenges of the project include:
1) developing the methods for identifying system and control parameters (including nonlinear control, such as Maximum Power Point Tracking command, or MPPT),
2) designing perturbations to obtain necessary measurements for identification, and,
3) developing cross-validation tests and identifying the limits of the produced models.
The challenges of the project include:
1) developing the methods for identifying system and control parameters (including nonlinear control, such as Maximum Power Point Tracking command, or MPPT),
2) designing perturbations to obtain necessary measurements for identification, and,
3) developing cross-validation tests and identifying the limits of the produced models.