Design Day<\/a>, an annual event showcasing students\u2019 solutions to real-world challenges.<\/p>\n\n\n\n
To identify the most cost-effective and sustainable microgrid setups, Gong first developed a simulation-based model that tested different combinations of energy sources and policy incentives. His goal was to find configurations that balanced affordability, reliability, and clean energy use.<\/p>\n\n\n\n
His simulations revealed that with well-designed policy incentives, microgrids could significantly increase their reliance on renewables while lowering overall system costs, without sacrificing performance and reliability. Systems that paired moderate carbon reduction incentives (like a small tax on fossil fuels) with strong renewable energy subsidies (such as rebates for installing solar panels) consistently outperformed others in both cost and emissions.<\/p>\n\n\n\n
To support those findings, he ran two advanced simulation models focused on solar and wind power: one predicted production using historical weather data, accounting for the ups and downs of natural variability, while the second used Markov Chain Monte Carlo (MCMC) to model the performance of key microgrid components\u2013such as solar panels and wind turbines\u2013perform over time, factoring in wear, failure risk, and maintenance needs.<\/p>\n\n\n\n
\u201cMicrogrids have to handle a lot of uncertainty,\u201d he said. \u201cWeather changes, machines break\u2014it all adds up. I wanted the model to reflect that.\u201d<\/p>\n\n\n\n
Next, Gong developed a tool that calculates total expenses, which considered a variety of financial factors, including the upfront investment needed to build and install a hybrid microgrid system, ongoing operating costs, and the economic impact of power outages and carbon emissions. He also factored in two key policy levers: incentives for using renewable energy and for reducing carbon emissions.<\/p>\n\n\n\n
With that foundation, he applied Mixed-Variable Simultaneous Perturbation Stochastic Algorithm (MSPSA), an optimization technique designed for complex, unpredictable systems). He compared its performance to Particle Swarm Optimization (PSO), a method used in energy planning to find the most efficient solutions by mimicking the collective behavior of birds or fish.<\/p>\n\n\n\n
\u201cWhat surprised me was how much better MSPSA performed compared to PSO\u2014not just in speed, but also in its ability to identify cleaner and more cost-effective solutions,\u201d he said. \u201cPSO isn\u2019t really built for noisy or unpredictable systems like those in renewable energy planning, yet it\u2019s still widely used in academic studies without much scrutiny. MSPSA, on the other hand, is specifically designed to handle randomness, which likely explains why it performed so much better in this context.\u201d<\/p>\n\n\n\n
<\/p>\n\n\n\n