As there are different types of energy storage systems (ESS), it is important to have a tool that determines the best alternative to be installed in the distribution systems. This study presents a methodology for optimal determination of size, type and site of ESS in distribution systems with distributed generation (DG). The methodology uses a genetic algorithm and a power flow calculated by the Electric Power Research Institute software OPENDSS. For the optimal ESS charge and discharge management uses a non‐linear optimisation model that has as objective an energy cost and loss reduction. In addition, the methodology makes an energy management of the system by limiting the reverse flow with a load shifting process and a cost penalty. Solar and wind power are considered as DG modelled in Homer. Then, a genetic algorithm is used to select the size, type and site of the ESS, considering different lifecycles and costs. This methodology is tested in the IEEE 123 nodes system simulated and adapted in the OPENDSS and the formulation is implemented in MATLAB via COM interface. Finally, results of the methodology are presented and discussed.

This paper proposes a flexible framework for scheduling and real time operation of electric vehicle charging stations (EVCS). The methodology applies a multi-objective evolutionary particle swarm optimization algorithm (EPSO) for electric vehicles (EVs) scheduling based on a day-ahead scenario. Then, real time operation is managed based on a rule-based (RB) approach. Two types of consumer were considered: EV owners with a day-ahead request for charging (scheduled consumers, SCh) and non-scheduling users (NSCh). EPSO has two main objectives: cost reduction and reduce overloading for high demand in grid. The EVCS has support by photovoltaic generation (PV), battery energy storage systems (BESS), and the distribution grid. The method allows the selection between three types of charging, distributing it according to EV demand. The model estimates SC remaining state of charge (SoC) for arriving to EVCS and then adjusts the actual difference by the RB. The results showed a profit for EVCS by the proposed technique. The proposed EPSO and RB have a fast solution to the problem that allows practical implementation.

This paper presents a new approach to the problem of defining an investment policy in battery energy storage systems in active distribution networks, taking into account a diversity of uncertainties. The proposed methodology allows the selection of type, capacity, and location of battery energy storage systems in distribution networks with distributed generation and electric vehicle charging stations. A mixed‐integer stochastic programming problem is cunningly approached with a metaheuristic, where fitness calculation with stochastic scenarios is performed by introducing an approximation to the operation costs in the form of a polynomial neural network, generated according to the Group Method of Data Handling—GMDH method, with strong computing speeding‐up. The quality of this approximation for heavy Monte Carlo simulations is assessed in a first case study using a 33‐bus distribution test system. The optimization planning model is then validated in the same test system using real data collected from solar and wind sources, demand, prices, and charging stations. Four types of batteries are compared considering degradation impact. The results demonstrate the practicality and advantages of this process.

This paper proposes a battery voltage model that is suitable for variable operation. The model combines the features of the Kinetic Battery Model (KiBaM) and voltage model (VM), and it improves the accuracy and quality of the solution, addressing four characteristics of operation: charging, discharging, rest after charge, and rest after discharge. This model will be known as 4-KiVM and shows low impact on computational burden. The proposed model can keep track of the voltage even when the load is inverted or turned off. To calibrate and validate the model, a NASA-provided dataset was used composed of a battery with variable charges and discharges, simulating real applications. A metaheuristic method based on tabu search is used to extract constants from this dataset and validate this hybrid model. In addition, a comparison of performance of the 4-KiVM against KiBaM, VM, and the electric circuit model (ECM) was made, showing its advantages. The results of the simulations showed a good prediction of the battery voltage response and SOC prediction in random (variable) use.