A dependable power supply is critical to the well-being of today’s society. According to forecasts, this issue will become increasingly relevant and pose numerous challenges that need to be faced. The energy demand is increasing. The load pattern might change and be more temperature-sensitive and volatile. The generation will be less predictable and dictated by geography, which may necessitate energy transfers over long distances. A liberalized power market aggravates the situation, and approvals for grid expansion are progressing very slowly. Therefore, it is becoming increasingly challenging for the transmission system operator to guarantee enough margins concerning stability and thermal overload. Power system flexibility is crucial for ensuring a stable future supply. System Protection Schemes, the subject of this dissertation, are a measure that can increase the grid’s flexibility. By selectively arming schemes, the system can operate with smaller reserves and be more effective (i.e., offering more capacity and lower risk). Control actions of such SPSs are, for example, generator rejection, load shedding, controlled separation and reconfiguration of the grid, etc. This thesis presents three adaptive arming methods based on mathematical models and different optimization procedures. The first approach uses a bi-level optimization technique for optimally arming multiple system protection schemes considering steady-state models. This procedure aims to minimize the amount of System Protection Schemes to be armed to detect all critical contingencies while keeping the number of possible trips due to non-critical contingencies as low as possible. This avoids triggering when unnecessary and positively influences the complexity and scheme’s security. One of the challenges is that the approach involves a mixed integer non-linear optimization problem, which is difficult and time-consuming to solve. The other two procedures presented in this thesis are event-based and predictive under-frequency load shedding, considering different dynamic models. The basic idea is to arm the protection concerning the current system state for a specific contingency, such as electric islanding. Thus, a fast scheme is proposed that requires no computation time during the post-contingency transient phase. The methods showed good results under different assumptions and could keep the quantities in the specified range. However, due to their feed-forward structure, uncertainties may be challenging, and the right granularity of the model is essential. In this work, the focus was on adaptive algorithms that are closely related to wide-area monitoring protection and control systems that depend on synchronized phasor measurement units. For this reason, various approaches to integrating Phasor Measurement Units into a real-time laboratory were also evaluated in the framework of this thesis. IEC 61850-based Phasor Measurement Units proved promising as they are a simple and costeffective method to realize a Wide-Area Monitoring System Hardware-inthe-Loop test platform.