Vineet Jagadeesan Nair

Lucas Pereira, Vineet Jagadeesan Nair, Bruno Dias et al.

We propose a comprehensive approach to increase the reliability and resilience of future power grids rich in distributed energy resources. Our distributed scheme combines federated learning-based attack detection with a local electricity market-based attack mitigation method. We validate the scheme by applying it to a real-world distribution grid rich in solar PV. Simulation results demonstrate that the approach is feasible and can successfully mitigate the grid impacts of cyber-physical attacks.

Vineet Jagadeesan Nair

A multiobjective, multiperiod global optimization framework is developed for the design, sizing, and dispatch of an islanded hybrid microgrid. System sizing is optimized over a one-year horizon and operational dispatch over a representative day, both at hourly resolution. The formulation minimizes lifecycle levelized cost of energy, emissions, lost load, and dumped energy, while maximizing renewable penetration. The approach identifies optimal capacities of renewable generation, storage, and backup generation that balance affordability, sustainability, reliability, and efficiency. Among the methods evaluated, particle swarm optimization is well suited for the nonconvex, multiobjective sizing problem. Results show that a solar PV-wind microgrid with lithium-ion battery storage and diesel backup consistently outperforms alternatives. Cost considerations dominate allocation among renewable sources, while sizing of renewables and storage is influenced by standby generation ratings due to reliability constraints. Pareto-optimal solutions reveal key tradeoffs among economic, environmental, and reliability objectives, showing that cost-only optimization can yield poorer emissions, reliability, and curtailment outcomes. Sensitivity analyses highlight the impact of fuel prices and storage costs on optimal design. Accurate sizing reduces unnecessary oversizing used to ensure reliability in off-grid systems, lowering upfront capital needs and improving affordability of clean electricity access. The dispatch model produces day-ahead schedules generally robust to short-term uncertainty, though disturbances increase reliance on fossil backup. Effective dispatch of batteries and backup generators is critical. The study also reviews microgrid design tools and methods, and addresses applications in sub-Saharan Africa.

Vineet Jagadeesan Nair, Morteza Vahid-Ghavidel, Anuradha M. Annaswamy

We show that dynamic coordination of distributed energy resources (DERs) can increase the capacity of low- and medium-voltage grids, improve reliability and power quality, and reduce solar curtailment. We develop three approaches to compute hosting capacity on a representative distribution grid with realistic scenarios. A deterministic iterative method provides insight into how dynamic operation and DER interactions enhance capacity and affect power flows, demonstrating clear gains over static methods even with low-to-moderate levels of storage and flexible demand. A stochastic programming approach jointly optimizes DER siting and sizing, showing that nodal colocation and complementary effects expand the feasible region of solar, heat pump, and battery penetrations by over 22X. This enables up to 200% solar, 100% battery, and 90% heat pump penetration. Batteries emerge as the most critical technology, followed by heat pumps and electric vehicles. A Monte Carlo-based extension shows that uncertainty significantly impacts hosting capacity and grid metrics, with 46% higher volatility under dynamic operation.

Vineet Jagadeesan Nair, Lucas Pereira

This proposal aims to develop more accurate federated learning (FL) methods with faster convergence properties and lower communication requirements, specifically for forecasting distributed energy resources (DER) such as renewables, energy storage, and loads in modern, low-carbon power grids. This will be achieved by (i) leveraging recently developed extensions of FL such as hierarchical and iterative clustering to improve performance with non-IID data, (ii) experimenting with different types of FL global models well-suited to time-series data, and (iii) incorporating domain-specific knowledge from power systems to build more general FL frameworks and architectures that can be applied to diverse types of DERs beyond just load forecasting, and with heterogeneous clients.