A novel approach of energy and reserve scheduling for hybrid power systems: Frequency security constraints

The increasing incorporation of solar and wind energy sources into the power system results in a diminishing power system inertia, which compromises the security of power system operation. This paper introduces a new scheduling method for hybrid power systems (HyPS), focusing on frequency security. It considers the timevarying characteristics related to frequency security requirements and the interdependence of components, such as generation units and the delivery network. Firstly, the optimal objective is to minimize overall operational costs and the pressure on energy delivery, ensuring the reserve capacities for fast frequency regulation remained within the predefined limits. Secondly, the frequency constrained unit commitment (FCUC) problem is formulated as a Markov decision process (MDP), and a deep reinforcement learning (DRL) framework is integrated to correct the advanced scheduling scheme in real-time. Thirdly, the coordinated scheduling scheme dynamically adjusts to operational events like fluctuations and line failures in the HyPS by employing the reward function of a comprehensive assessment. Finally, numerical results demonstrate the practicality of the novel coordinated scheduling approach for hybrid power systems through simulations in the Grid2Op environment. As a result, the frequency security margin of the HyPS is capable of increasing the daily average level of the frequency security margin by 3% across diverse scenarios, including fluctuations in power demand, renewable energy variations, and power line failures. PE-interfaced renewable energy sources are thereby able to integrate with conventional synchronous units as a hybrid power system to support frequency regulation services.