Cyber- Physical Robustness Enhancement Strategies for Demand Side Energy Systems

Abstract

An integrated Cyber-Physical System (CPS) system realizes the two-way communication between end-users and power generation in which customers are able to actively re-shaped their consumption profiles to facilitate the energy efficiency of the grid. However, large-scale implementations of distributed assets and advanced communication infrastructures also increase the risks of grid operation. This thesis aims to enhance the robustness of the entire demand-side system in a cyber-physical environment and develop comprehensive strategies about outage energy management (i.e., community-level scheduling and appliance-level energy management), communications infrastructure development, and cybersecurity controls that encounter virus attacks. All these aspects facilitate the demand-side system’s self-serve capability and operational robustness under extreme conditions and dangerous scenarios.

The research that contributes to this thesis is grouped around and builds a general scheme to enhance the robustness of CPS demand-side energy system with outage considerations, communication network layouts, and virus intrusions.

Under system outage, there are two layers for maximizing the duration of self-power supply duration in extreme conditions. The study first proposed a resilient energy management system for residential communities (CEMS), by scheduling and coordinating the battery energy storage system and energy consumption of houses/units. Moreover, it also proposed a hierarchical resilient energy management system (EMS) by fully considering the appliance-level local scheduling. The method also takes into account customer satisfaction and lifestyle preferences in order to form the optimal outcome.

To further enhance the robustness of the CPS system, a complex multi-hop wireless remote metering network model for communication layout on the CPS demand side was proposed. This decreased the number and locations of data centers on the demand side and reduced the security risk of communication and the infrastructure cost of the smart grid for residential energy management. A novel evolutionary aggregation algorithm (EAA) was proposed to obtain the minimum number and locations of the local data centers required to fulfill the connectivity of the smart meters.

Finally, the potential for virus attacks has also been studied as well. A trade-off strategy to confront viruses in the system with numerous network nodes is proposed. The allocation of antivirus programs and schemes are studied to avoid system crashes and achieve the minimum potential damages. A DOWNHILL-TRADE OFF algorithm is proposed to address an appropriate allocation strategy under the time evolution of the expected state of the network. Simulations are conducted using the data from the Smart Grid, Smart City national demonstration project trials.