Skip to main content
Infrastructure Resilience Conference 2018

Full Program »

A System Approach for Feedback Control Scheduling in Contingency and Emergency Plans for Power Grid Restoration

To date, contingency and emergency plans for geographical distributed systems, such as the power grid, gas or water system, as well as the communication system, are developed in advance. The list of actions for those plans is hereby usually determined by the solution to optimisation problems with capacity and time constraints to achieve a certain quality of service within a certain time. After a disruptive event, some disturbances affecting the system state become highly unpredictable. The achieved result by applying the fixed plan may differ from the optimal solution or it could even fail, leading to a system collapse resulting in a low system resilience. In view of this, this paper first shows that the development of a fixed plan in advance represents an open loop controller in terms of control engineering, while it also states the corresponding pitfalls. Building upon the principles of control theory, this paper addresses the question whether higher resilience in terms of recovery time to a given measurement of performance is achieved by implementing a closed loop strategy instead. To answer this, we are developing a model for closed loop contingency and emergency planning and introduce a case study on power grid restoration to demonstrate preliminary results. In this study, we take human performance of the restoration teams into account as well as geographical accessibility constraints of the physical power grid infrastructure. We show that one can view the system describing the plan execution as a control loop. Thus, we are employing the methodology of mobile model predictive control (MoMPC) which has been introduced for manual human control of irrigation canals to enable closed loop feedback policies for contingency and/or emergency planning with constraints. In this approach, the control inputs and measurements are applied through human actuators on a graph, representing the geographical distribution of the system. The assignment of restoration teams is re-scheduled at each time step based on the current system state of the power grid to carry out inspection and restoration tasks. Our study contributes to existing literature by thoroughly discussing the model structure as a control system, its components, such as the physical infrastructure of the power grid and its geographical distribution, as well as the human factors aspects, and how all these shall be treated in a joint system. Future work will enhance the sub-system structure with more refined models of human performance, the geographical integration of the power grid infrastructure as well as its extraction from available data sources.

Felix Kottmann
ETH Zurich, Future Resilient Systems, Singapore - ETH Centre
Singapore

Miltos Kyriakidis
ETH Zurich, Future Resilient Systems, Singapore - ETH Centre
Singapore

Vinh Dang
Laboratory for Energy Systems Analysis, Paul Scherrer Institute
Switzerland

Giovanni Sansavini
ETH Zurich, Reliability and Risk Engineering Group
Switzerland

 

Powered by OpenConf®
Copyright ©2002-2016 Zakon Group LLC