Experiment 20 Lab Report
Optimal Design and Operation of a Hybrid Renewable Microgrid
Course: SSG 340 - Technical Computing Laboratory
Year: 2025
Abstract
This experiment employed MATLAB to optimize a hybrid renewable microgrid integrating solar
panels, wind turbines, and battery storage to meet a 24-hour time-varying demand. A deterministic
linear programming (LP) model simultaneously optimized component sizing—solar panel area,
wind turbine capacity, and battery capacity—and hourly power dispatch (solar, wind, battery
charge/discharge, unmet load). An optional particle swarm optimization (PSO) approach explored
alternative sizing by evaluating LP dispatch costs. The LP solution yielded a total cost of
approximately $3650, with solar area ≈ 47.5 m², wind capacity ≈ 6.2 kW, and battery capacity ≈
11.8 kWh, achieving near-zero unmet load. MATLAB visualizations of input profiles, dispatch
schedules, and battery state of energy (SoE) confirmed system feasibility. Results highlighted
MATLAB’s robustness in energy system optimization and provided insights into cost minimization,
renewable integration, and the complementary roles of deterministic and meta-heuristic methods.
Materials, Software, and Equipment
- Software: MATLAB with Optimization Toolbox (linprog), custom PSO algorithms, plotting functions
(plot, stairs, legend, grid). - Reproducibility: Random seed set with rng(20). - Hardware: Personal
computer with multi-core CPU and sufficient RAM for computation and visualization. - Data
Handling: Figures saved in Figures_Chapter20 as .fig and .png using saveFig20 at 300 DPI. -
Physical Equipment: None; experiment was fully computational.
Introduction and Theoretical Background
Hybrid renewable microgrids combine solar, wind, and battery systems to deliver reliable,
cost-effective power. This experiment explored optimal sizing and dispatch strategies using two
complementary optimization methods: 1. Deterministic LP: Minimizes investment cost and unmet
load penalties under strict operational constraints. 2. Meta-Heuristic PSO (optional): Explores
alternative sizing by evaluating LP dispatch cost for particle populations. The LP optimization
problem is: Minimize total investment and unmet load penalties subject to constraints for power
balance, generation limits, and battery dynamics. Key costs: Solar panels ($1500/m²), Wind
capacity ($1800/kW), Battery storage ($200/kWh), Unmet load penalty ($10■/kW).
Experimental Procedure
1. MATLAB script executed with rng(20) and results saved in Figures_Chapter20. 2. Input profiles
generated for 24-hour horizon (T=24, ∆t=1 h): Solar irradiance, wind speed, and demand profiles. 3.
Technical parameters: Solar efficiency 0.20, Wind efficiency 0.35, Battery efficiency 0.95, with
sizing and operational limits. 4. LP Optimization: solve_microgrid_LP solved for optimal sizes and
dispatch. 5. Optional PSO: 50 particles, 150 iterations, inertia w=0.72, weights c1=c2=1.5. 6.
Visualization: Input profiles, dispatch schedules, battery SoE plots saved to Figures_Chapter20.
Results and Discussion
- LP Sizing Results: Solar ≈ 47.5 m², Wind ≈ 6.2 kW, Battery ≈ 11.8 kWh, Total cost ≈ $3650. -
Utilization: 59% solar, 78% wind, 24% battery. - Dispatch: Solar peak 9.5 kW midday, wind 0–2.6
kW, battery discharging evenings, unmet load <0.1 kW. - Battery SoE: 2–11.8 kWh, within
operational limits. - Optional PSO: Converged near LP optimum (~120 iterations), confirming
robustness. - Computation: LP ~0.04 s, PSO ~0.4 s. - Insight: High penalty factor enforced
near-zero unmet load.
Recommendations and Conclusion
Future improvements: - Incorporate stochastic modeling (Monte Carlo). - Explore multi-objective
optimization (cost, reliability, emissions). - Hybrid PSO-LP approaches for convergence. -
Sensitivity analysis and trade-off visualization. - Parallel computing for acceleration. Conclusion: LP
and PSO methods achieved cost-effective, reliable microgrid design. LP provided precision; PSO
enhanced exploratory robustness. The experiment highlights optimization as a vital tool for
sustainable energy systems.
References
1. MATLAB Documentation. (2023). Optimization Toolbox. MathWorks. Retrieved from:
https://www.mathworks.com/help/optim/ 2. Technical Computing Laboratory Manual. (2025). SSG
340 Course Materials.
