Contact Person:
Christian Hans
Our research in this area employs control methods for real-world applications in the domain of complex electrical power systems with a large share of decentralized, uncertain renewable generation. The overall objective is to enable carbon-free energy systems using automatic control. In what follows these applications are highlighted. They include distributed and decentralized low-level control as well as optimization-based high-level control, and control methods for wind power plants.
In future electrical power systems, conventional generators, such as coal power plants, will be mainly replaced by inverter-interfaced renewable sources and storage units. Our research in this field concerns different dimensions of low-layer control of such systems that comprise conventional and renewable generators as well as inverter-interfaced storage units. Modeling issues, which are a cornerstone for most control analysis and synthesis problems described below, were investigated in [Schiffer et al., 2016b].
Droop control, a simple decentralized scheme, has been a popular low-level control approach. Within the Control Systems Group, conditions for stability and power sharing under droop control were derived in [Schiffer et al., 2014a,b] for inverter-based microgrids and in [Schiffer et al., 2013] for grids with mixed rotational and electronic-interfaced units. In [Schiffer et al., 2015a, 2017], the analysis was extended to models that consider inaccurate clocks, induced by the fact that each individual unit is operated with its own digital processor. The effect of delays in grids with distributed rotational and electronic generation under droop control was thoroughly investigated in [Schiffer et al., 2016a].
The aforementioned analysis showed that droop control is not suitable to achieve reactive power sharing and typically leads to steady-state frequency deviations. As a consequence, in [Schiffer et al., 2015b, Schiffer. et al., 2014, Krishna et al., 2019] consensus-based distributed voltage control laws which guarantee reactive power sharing were proposed. Moreover, in [Krishna et al., 2017] different distributed secondary control schemes were thoroughly compared with respect to their steady-state frequency error, and in [Krishna et al., 2018], a consensus-based control law for accurate frequency restoration has been proposed. Furthermore, in [Krishna et al., 2020], robustness and steady-state performance of distributed secondary frequency control schemes have been investigated.
Microgrids are small power systems that can be operated connected to or isolated from a larger grid. They typically comprise storage units, renewable and conventional generators as well as loads. “Operation control” refers to the top, or tertiary, control level within a hierarchical control scheme and is typically based on comparatively simple models of the plant under low-level control. One central research question concerning microgrids is: how to operate microgrids with very high renewable share, i.e., how to control the energy of storage units, and how to maximize infeed from uncertain renewable sources without compromising a safe operation?
To answer this question, various model predictive control (MPC) schemes were derived within the Control Systems Group. They can be distinguished by the way they handle the uncertain load and renewable infeed: (i) certainty equivalence MPC [Hans et al., 2014], where a nominal forecast is fully trusted; (ii) minimax MPC [Hans et al., 2014, 2015a], where time-varying forecast intervals are assumed; (iii) risk-neutral stochastic MPC [Hans et al., 2015b], where a forecast probability distribution is fully trusted; and (iv) risk-averse MPC [Hans et al., 2020], where a forecast probability distribution is not fully trusted. The different approaches were thoroughly compared in different case studies in [Hans, 2021]. Moreover, in [Löser et al., 2019], fallback strategies for operation control of microgrids with communication failures were proposed. Additionally, in [Strenge et al., 2020b,a], learning based strategies for operation control were investigated. In [Strenge et al., 2019], a global controller is synthesized to improve the performance of a microgrid formed by interconnected prosumers.
One approach to handle the operation of complex future energy systems with a large share of small-scale renewable energy sources is to partition the overall grid into smaller microgrids. To the outside world, each microgrid appears as a single entity that can provide or consume power. Allowing power exchange between these entities typically increases the overall system performance by exploiting effects like smoothing by geographical dispersion of renewable generators. One important research question in this area is: how to enable optimal trading among a network of interconnected microgrids while preserving the autonomy of each microgrid? Different distributed MPC schemes were derived within the Control Systems Group to answer this question. In [Hans et al., 2019], a hierarchical distributed MPC approach was presented. The approach considers a central coordinator that takes care of the transmission network which connects the individual microgrids. Power setpoints are found by consecutively solving (and communicating the results of) local optimization problems and a problem at the central coordinator. In [Sampathirao et al., 2021], the approach was extended by adding a condition that preserves the self-interests of each microgrid. Moreover, the central coordinator could be removed and the algorithm could be implemented in a fully distributed way that only requires peer-to-peer communication between neighboring microgrids.
Many wind farms are operated to a considerable extent in so-called power tracking mode in order to meet requirements of power system operators. In this mode, wind energy converters do not feed in the uncertain weather-dependent maximum possible power, but are operated with reduced power. Within current research activities at the Control Systems Group, existing degrees of freedom in the design of wind energy converters and wind farm controllers are exploited in order to achieve a better operation of the overall wind farm in power tracking mode.
Academia
Industry