Improving the stability of the electrical grid with the penetration of renewable energies

Abstract

This work is part of a study on the optimal integration of decentralized generation (DG) based on renewable photovoltaic energy sources and capacitor banks in radial distribution networks, particularly the widely used IEEE 33-JB, IEEE 69-JB, and ALG-AB-Hassi Sida 157 networks in the literature for planning and operating distribution systems. To optimize the use of photovoltaic sources and capacitor banks, recent meta-heuristic optimization methods have been proposed to determine their optimal location and size. These methods include the Grey Wolf Optimizer (GWO), Particle Swarm Optimization (PSO), Whale Optimization Algorithm (WOA), and the EM-BT algorithm. However, the "backward-forward" method has been used for power flow calculations.

For optimal integration, three studies have been proposed: the first aims solely to minimize energy losses in distribution networks by integrating multiple decentralized generation sources. The second study focuses on minimizing daily energy losses and voltage profile deviations, adapting to dynamic load conditions over 24 hours. This goal is captured in a unified objective function (Obj1), which extends to minimizing costs related to energy losses and energy supplied by photovoltaic sources and reactive power source capacitor banks (PVRES-CBs), thereby forming an overall objective model (Obj2).

The efficiency of the proposed MOMVO algorithm is rigorously evaluated through a comparative analysis with three other optimization algorithms: the Multi-Objective Jellyfish Search (MOJS), the Multi-Objective Floral Pollination Algorithm (MOFPA), and the Multi-Objective Lichtenberg Algorithm (MOLA). The comparison produces a Pareto front, which is essential for decision-makers or operators of electrical systems facing ambiguous or fuzzy objectives for each function. Fuzzy logic theory is applied to help select the optimal operating point from the set of Pareto optimal solutions by identifying the best non-dominated solution while maximizing the normalized sum of the membership function values across all objectives.

Moreover, the study explores variations in the best compromise solutions obtained from the mentioned algorithms, providing valuable insights into their performance and suitability under different network conditions. Another study was conducted using the Energy Valley Optimizer, which demonstrated its effectiveness compared to other optimization algorithms, such as the Liver Cancer Algorithm, the Morse Optimization Algorithm, and the Zebra Optimization Algorithm. This algorithm optimizes the overall cost objective function of energy over 24 years by integrating photovoltaic sources and capacitor banks, taking into account the intermittent nature of these sources and load variations over 24 hours. The results show high-quality solutions compared to other methods reported in the literature. The simulation results also indicate that the proposed algorithms are reliable and applicable for the optimal integration of decentralized generation sources in various distribution networks of different sizes and complexities.