Nonlinear consensus for improved resilience of distributed secondary frequency control

The integration of renewable energy sources into electrical power systems requires an increasing degree of automation in grid control and thereby leads to a higher vulnerability with respect to cyber-attacks. The response to this challenge should be twofold: first, it should be made as difficult as possible to perform successful cyber-attacks through physical and IT security measures. Second, if cyber-attacks are successful anyway, the attacked control system should be maximally resilient against the resulting manipulations. In this paper the latter approach is considered. We introduce a notion of trust into a linear consensus scheme for secondary frequency control in power systems. We show the resulting nonlinear consensus to preserve important properties like stability and optimality of the linear consensus concept, while achieving a higher level of resilience against cyber-attacks.