The change of course of energy resources in electrical grids to renewable energies necessitates the integration of monitoring and control devices in the grids. Among different layers of the electrical grid, low-voltage distribution grids are moving from a passive structure to an active system, meaning that monitoring, communication, and control devices are used to cope with changes in load consumption profiles and renewable energy uncertainties.
 
The integration of the mentioned control, monitoring, and communication parts changes the nature of the system from a physical system to a combined cyber and physical system, which falls under the category of cyber-physical systems. Due to the interconnections and interactions between the cyber and physical components, they mutually influence one another. For instance, delays in the communication network may lead to loss of performance and even instability. Therefore, it is crucial to analyze the integrated system and deal with the mutual effects of subsystems.
 
The main negative impacts of the cyber subsystem on the physical system are communication imperfections. Transmission intervals, delays, and packet losses are the main properties that can affect the physical system. The transmission interval is especially important for low-voltage distribution grids, since communication intervals are long due to the high number of consumers. Additionally, the threat of cyber-attacks in communication network is another crucial factor that can affect the physical system.
 
Therefore, the main focus of this thesis is to model, analyze, and control the cyber-physical power systems in the presence of communication imperfections and threads. The focus especially is on voltage control in low-voltage distribution grids and load frequency control.
 
Given that both the cyber and physical subsystems play a crucial role in this process, it is essential to model the system such that it contains the necessary properties of both subsystems. In other words, a cyber-physical model is of interest in analyzing and controlling the system.
 
It is investigated that hybrid models are appropriate for modeling cyber-physical systems, given the presence of both discrete (cyber subsystem) and continuous (physical subsystem) dynamics. The presented models can represent necessary features of the cyber subsystem along with physical system dynamics.
 
The voltage control problem in low-voltage distribution grids has been studied in this thesis. A centralized control structure has been proposed. To meet the requirements of the grid, invariance-based control is used. However, the invariance control problem has been updated due to the special conditions of the system. The method is able to keep the voltages in the desired ranges, which meets the system requirements.
 
Load frequency control in power systems is used to investigate the effects, mitigate, and detect cyber-attacks. Denial-of-service (DoS) attacks, as the most common cyber-attack, and false data injection are studied.
 
Since the DoS attacks affect the communication properties (packet losses, delays, and communication intervals), system stability and performance considering communication properties are discussed. To this end, a hybrid system approach is utilized to model the system and analyze the system. As a consequence, bounds and trade-offs for communication system properties are investigated to ensure system stability and keep the system at the desired performance.
 
Finally, a robust method is used to mitigate and detect false data injection attacks in the load frequency control loop. The method is based on a single module; however, in order to avoid conservatism in the design of either controller and detector, the problem criteria are divided into two problems. The results show that the control and detection processes are able to regulate frequency in the presence of disturbances, detect attacks, and control the system in the presence of undetected attacks.

Reception 
After the defence there will be a small reception at Fredrik Bajers Vej 7A4-106