Rishav Sen

PhD

PhD Student

advanced data visualization agent-based models Bayesian Modeling Deep Learning Mathematical modelling Python Pytorch reinforcement learning Simulations Statistical Modeling Time Series Analysis

Peer-reviewed publication(s) 1st author peer-reviewed publication(s) Conference Poster(s) Reviewed Publication(s)

Designed a hierarchical control framework combining mixed-integer linear programming and multi-agent reinforcement learning to coordinate distributed energy assets in a vehicle-to-building charging system. Validated the framework on real field data, demonstrating over 10% cost reduction compared to state-of-the-art baselines. Also built a negotiation framework to incorporate user flexibility under uncertainty, and co-developed a discrete event simulator for benchmarking demand-response strategies.

Led optimization modeling for electrifying a mixed public transit bus fleet. Developed a hierarchical formulation jointly solving trip assignment and charge scheduling, reducing operational costs by 6% and cutting greenhouse gas emissions by 10 tons. Built open-source simulation tools (E-Transit-Bench, BTE-Sim) now used by academic partners for city-scale transit resilience studies.

Developed a Python-based open-source data pipeline fusing US Census and traffic data to synthesize origin-destination commute flows for city planners. Separately, built a spatial optimization model using crash data to identify optimal emergency hub placements, significantly reducing projected ambulance response times across a metropolitan area.