Victor D. Dorobantu

Papers

Total Citations

H-Index

About

Victor D. Dorobantu is a roboticist advancing the frontier of dynamic legged locomotion through the principled application of nonlinear control and optimization. His research centers on achieving dynamically stable behaviors in complex robotic systems, with a particular focus on the formidable challenge of controlled hopping. Dorobantu’s most prominent work, "Nonlinear Model Predictive Control of a 3D Hopping Robot: Leveraging Lie Group Integrators for Dynamically Stable Behaviors" (2023, 14 citations), directly addresses this problem by employing geometric numerical integration to manage the robot’s extended under-actuation and brief, critical ground contact phases. This work provides a rigorous framework for modulating ground interactions to regulate global state, a key step toward more agile and resilient legged machines. Beyond application, Dorobantu also contributes to the theoretical foundations of reinforcement learning, as seen in his work on "Compactly Restrictable Metric Policy Optimization Problems" (2022), where he establishes well-posedness conditions for policy optimization in continuous state and action spaces. By bridging rigorous control theory with practical robotic implementation, Dorobantu’s research offers a compelling blueprint for achieving robust, high-performance locomotion in the real world.