Measurement system and technique for future active distribution grids
- Messsystem und Messtechnik für zukünftige aktive Verteilnetze
Liu, Junqi; Monti, Antonello (Thesis advisor)
1. Aufl.. - Aachen : E.ON Energy Research Center, RWTH Aachen University (2014)
Dissertation / PhD Thesis
In: E.On Energy Research Center 18 : ACS, automation of complex power systems
Page(s)/Article-Nr.: XII, 144 S. : Ill., graph. Darst.
Zugl.: Aachen, Techn. Hochsch., Diss., 2013
The changes in the new landscape of active distribution grids are accompanied by increasing system complexity, dynamics and uncertainty that also raise new challenges and issues for system reliability and stability. A real-time measurement and monitoring system providing accurate and extensive measurement information is needed as the prerequisite for enabling reliable and efficient grid control and operation. However, an appropriate measurement and monitoring system is still not established in distribution grids. Therefore, this dissertation addresses three aspects concerning measurement system planning, system operational parameter estimation and system states tracking with the overall objective to design and develop novel real-time measurement system and techniques for future active distribution grids. The design of an overall measurement infrastructure is first addressed by developing a meter placement optimization procedure which builds the fundamental for facilitating future measurement systems. A design procedure is proposed for distribution system state estimation, where heterogeneous measurements including traditional and emerging measurement technologies are combined. The design problem is posed in terms of a stochastic optimization with the goal of guaranteeing the overall accuracy of the state estimation while minimizing the investment costs. The robustness of the meter placement in case of measurement device failure and degradation as well as the lack of measurement information with non-Gaussian probabilistic nature is also taken into account. The proposed method can assist distribution system operators in making investment decisions to upgrade the grid for the upcoming changes in the active distribution grids. As next aspect, an accurate measurement technique for system operational parameters in the grids is investigated and a new approach for instantaneous phasor and frequency estimation is proposed. A modified Taylor-Kalman filter is developed for accurate dynamic phasor estimation under dynamic and distorted conditions. In addition, accurate estimation of frequency and rate of change of frequency is designed using weighted smoothing differentiation based estimator in combination with the proposed dynamic phasor estimator. The proposed technique provides an alternative method for phasor, frequency information and even power quality measurements with high accuracy for future active distribution grids. Considering the interconnected nature and increasing dynamics of the system components and the grids, a fully decentralized Kalman filtering approach is proposed to facilitate system level awareness. This decentralized approach is enabled by introducing the virtual disturbance concept for power system decoupling. Accurate tracking of local dynamic states and interactions among the subsystems is realized by designing augmented local Kalman filter with proper virtual disturbance modeling without any communication effort. This enables gaining local and to some extent system level awareness at the same time. The herein presented approach is well suited for the agent-based decentralized control to enable "plug&play" features.