Enhancing the Reliability of Residential Microgrids Through Genetic and Ant Colony Algorithms

## Assessment and Optimization of Residential Microgrid Reliability Using Genetic and Ant Colony Algorithms

Abstract

The reliability of sustainable energy systems is influenced by the variability of renewable energy sources, storage limitations, and fluctuations in residential demand, which can lead to energy deficits and service interruptions. This study aims to diagnose and optimize the reliability of a residential microgrid powered by photovoltaic (PV) and wind energy, along with battery energy storage systems (BESSs). We employ genetic algorithms (GAs) and ant colony optimization (ACO) to evaluate system performance using metrics such as loss of load probability (LOLP), loss of supply probability (LPSP), and availability. The test system comprises a 3.25 kW PV system, a 1 kW wind turbine, and a 3 kWh battery. Simulations conducted with real consumption, solar irradiation, and wind speed data demonstrate that initial reliability is limited, with an availability of 77% and high LOLP (22.7%) and LPSP (26.6%). Optimization through metaheuristic algorithms notably improved these indicators, reducing LOLP to 11% and LPSP to 16.4%, while increasing availability to 89%. The optimization also achieved a better balance between generation and consumption, particularly during low-demand periods, with ACO enhancing the distribution of wind and PV generation. This study concludes that employing metaheuristics effectively enhances the reliability and efficiency of autonomous microgrids, optimizing energy balance and operational costs.

1. Introduction

The current energy crisis highlights the urgent need to transition from fossil fuels to a sustainable and decarbonized energy system. The 2030 Agenda emphasizes the necessity to harness non-polluting renewable energy sources while improving efficiency and generation capacity. In this context, microgrids have gained prominence as effective solutions to promote renewable energy consumption and enhance electrical system reliability. A microgrid is a localized energy system that integrates distributed generation, energy storage, controllable loads, and microturbines, capable of operating either in conjunction with the main grid or independently.

Photovoltaic and wind power have become essential in power generation due to their abundance and recent technological advancements. However, their intermittent nature poses challenges to supply stability. To mitigate this variability, energy storage systems (ESSs), particularly BESSs, have been proposed to balance electricity supply and demand, enhancing the reliability and sustainability of electrical systems.

Residential microgrids efficiently integrate renewable energy generation with battery storage, either operating independently or connecting to existing power grids to increase resilience and facilitate a more sustainable energy model. Despite their advantages, the inherent variability of renewable sources, limited storage capacity, and demand fluctuations continue to challenge the reliability of these sustainable energy systems, leading to energy deficits and affecting supply availability.

A primary challenge in microgrid design and operation is achieving an optimal balance between energy generation, storage, and consumption. This involves minimizing LOLP and optimizing implementation and operational costs. Continuous diagnosis and optimization are essential due to factors affecting microgrid performance, including changes in energy demand patterns, fluctuations in renewable generation, and storage system degradation.

Traditional microgrid sizing techniques, such as the peak load method, average load method, energy balance method, and reliability analysis method, each have their pros and cons. However, many optimization algorithms exist, categorized into deterministic, metaheuristic, and AI-based methods. Deterministic algorithms provide exact solutions but can be computationally expensive for complex problems. Metaheuristic algorithms like GA and ACO are effective in power system optimization due to their ability to address complex, nonlinear issues.

This research aims to diagnose and optimize the reliability of a residential microgrid using GA and ACO while evaluating performance based on reliability metrics and costs. The study addresses the following gaps:
– The lack of integration of reliability metrics in the optimal sizing and configuration of hybrid microgrids.
– Limitations of AI-based optimization algorithms due to their dependence on large training datasets.
– Existing methods’ inability to adapt to real-time variations in demand and generation.
– Limited validation of proposed optimization techniques with real data.

2. Materials and Methods

The methodology for modeling the residential microgrid includes evaluating PV and wind generation, modeling the BESS, assessing power balance, and analyzing reliability metrics. The optimization is performed using GA and ACO with defined fitness functions, operational constraints, and convergence criteria.

2.1 Renewable Energy Generation Modeling

Real data on estimated residential demand is combined with environmental variables such as solar irradiance and wind speed to model energy generation. For PV generation, the power produced is calculated using the effective area of panels, system efficiency, and total number of solar panels. Wind power generation is modeled based on wind speed and turbine characteristics.

2.2 Storage System Modeling (BESS)

The state of charge (SOC) indicates the energy stored in the battery and is calculated based on charge and discharge efficiencies. Operating limits ensure the battery does not fall below a minimum SOC to avoid damage.

2.3 Power Balance Assessment

The overall energy demand is assessed to ensure a stable electricity supply. Energy not supplied (ENS) is calculated when generation does not meet demand, and reliability metrics like LOLP and LPSP are established.

2.4 Reliability Metrics

LOLP, LPSP, and availability are key indicators of microgrid reliability. LOLP measures the probability of energy shortages, while LPSP indicates unmet energy demand. Availability reflects the proportion of time the system can meet demand.

2.5 Optimization Simulation

The objective is to optimize the design of the microgrid to minimize energy deficits and costs while maximizing reliability and efficiency. The fitness function evaluates system performance, considering various factors.

3. Results

3.1 Microgrid Evaluation and Diagnosis

The initial configurations of the microgrid reveal significant limitations in reliability, with an availability of 77% and high LOLP and LPSP values. The analysis of daily SOC and energy production indicates a need for optimization.

3.2 Optimization of the Microgrid

The optimization process resulted in improved reliability and cost metrics. GA optimization led to an increase in PV capacity and battery size, reducing LOLP and LPSP and improving system availability. ACO optimization achieved similar reliability improvements at a lower cost. The configuration obtained through exhaustive search provided the highest reliability but at a significantly higher cost.

4. Discussion

The findings indicate that initial configurations of the microgrid struggle with power continuity, necessitating optimization. GA and ACO effectively enhance reliability while balancing costs. The study suggests that optimization strategies are crucial for improving microgrid performance, especially in fluctuating demand scenarios.

5. Conclusions

This research demonstrates that GA and ACO are effective tools for assessing and optimizing the reliability of residential microgrids. The optimized configurations significantly improved availability and reduced the likelihood of energy shortages, showcasing the potential for these algorithms to enhance the stability and economic viability of autonomous microgrids.