The computational burden and the time required to train a deep reinforcement learning (DRL) can be appreciable, especially for the particular case of a DRL control used for frequency control of multi-electrical energy storage (MEESS). This paper presents an assessment of four training configurations of the actor and critic network to determine the configuration training that produces the lower computational time, considering the specific case of frequency control of MEESS. The training configuration cases are defined considering two processing units: CPU and GPU and are evaluated considering serial and parallel computing using MATLAB® 2020b Parallel Computing Toolbox. The agent used for this assessment is the Deep Deterministic Policy Gradient (DDPG) agent. The environment represents the dynamic model to provide enhanced frequency response to the power system by controlling the state of charge of energy storage systems. Simulation results demonstrated that the best configuration to reduce the computational time is training both actor and critic network on CPU using parallel computing.@en

This research paper presents a method that uses measurements of voltages angles, as provided by phasor measurement units (PMU), to accurately detect the sudden disconnection of a generation unit from a power grid. Results in this research paper have demonstrated, in a practical fashion, that a multi-layer perceptron (MLP) neural network (NN) can be appropriately trained to detect and identify the sudden disconnection of a generation unit in a multi-synchronous generation unit power system. Synthetic time-series bus voltage angles considering low inertia scenarios in the IEEE 39 bus system were used to train the MLP NN. The training process is speeded up by using four GPUs hardware. The simulations results have confirmed the successful detection and identification of the generator outage.@en