(369a) Robust Fault-Tolerant Control of Distributed Energy Systems

Sun, Y., University of California, Davis
El-Farra, N. H., University of California, Davis

Distributed Energy Resources (DERs) are a suite of on-site, grid-connected or stand-alone technology systems that can be integrated into residential, commercial, or institutional buildings and/or industrial facilities. These energy systems include distributed and hybrid generation technologies, renewable energy sources, energy storage, thermally activated technologies that use recoverable heat for cooling, heating, or power; transmission and delivery mechanisms; control and communication technologies; and demand-side energy management tools. Such distributed resources offer advantages over conventional grid electricity by offering end users a diversified fuel supply; higher power reliability, quality, and efficiency; lower emissions and greater flexibility to respond to changing energy needs.

Several efforts have been made over the past decade towards the development and implementation of control strategies for DERs. Examples include the use of conventional and model-based feedback control systems to regulate various types of grid-connected DERs in order to enhance power system stability (e.g., [1]), mitigate power quality problems (e.g., [2]) and improve the continuity of electricity supply, the development of various distributed control and coordination architectures using multi-agent system approaches (e.g., [3]), and the design of networked control systems to regulate DERs over communication networks [4]. While the aim of these studies has been mainly to demonstrate the feasibility of the developed control strategies, an important problem that has received less attention and remains to be addressed -- both in the formulation and solution of DER control problems -- is the integration of fault detection and handling capabilities in the control system design. This is an important problem given the fact that the distributed power market is primarily driven by the need for super-reliable, high-quality power and the fact that the impact of local faults and malfunctions in the distributed power network can be quite substantial, especially when the DERs are integrated into grid operations and the disruptions in power supply caused by local failures in the DERs has the potential to cascade through the network and interfere with grid operations.

Motivated by these considerations, we present in this contribution a hierarchical structure for robust fault detection and reconfigurable control of distributed energy resources subject to external disturbances and control actuator faults. The structure consists of distributed monitoring and fault-tolerant control systems that perform fault detection and control system reconfiguration at the local level, together with a supervisor that communicates with the local controllers and provides high-level oversight and contingency measures in the event that local fault recovery is not possible. Initially, an observer-based output feedback controller is designed for each DER to regulate its power output at the desired set-point in the absence of faults. The design accounts explicitly for practical implementation issues such as measurement sampling and plant-model mismatch. To ensure robust fault detection, the observer is designed using unknown input observer design techniques to decouple the effect of external disturbances on the state estimation error which is used as a residual for fault detection. An inter-sample model predictor is embedded within the local control systems and used to provide estimates of the output measurements between sampling times. Fault detection is performed by comparing the output of the observer with that of the DER at the sampling times, and using the discrepancy as a residual. An explicit characterization of the minimum allowable sampling rate that guarantees both closed-loop stability and residual convergence in the absence of faults is obtained and used as the basis for deriving (1) a time-varying alarm threshold on the residual which can be used to detect faults for a given sampling period, and (2) a controller reconfiguration law that determines the feasible fall-back control configurations that can be used to preserve stability and minimize performance deterioration. The design and implementation of the fault detection and fault-tolerant control architecture are finally demonstrated using an array of solid oxide fuel cell plants in a power distribution network.


[1] Sedghisigarchi, K. and A. Feliachi, ``Dynamic and transient analysis of power distribution systems with fuel cells-part II: control and stability enhancement," IEEE Transactions on Energy Conversion, 19:429-434, 2004.

[2] Marei, M., E. El-Saadany, and M. Salama, ``A novel control algorithm for the DG interface to mitigate power quality problems," IEEE Transactions on Power Delivery, 19:1384-1392, 2004.

[3] Dimeas, A. and N. Hatziargyriou, ``Operation of a multiagent system for microgrid control.," IEEE Transactions on Power Systems, 20:1447-1455, 2005.

[4] Sun, Y., S. Ghantasala, and N. H. El-Farra, ``Networked control of distributed energy resources: Application to solid oxide fuel cells," Ind. Eng. Chem. Res., 48:9590-9602, 2009.