Mircea Lazar

Papers

Total Citations

H-Index

About

Mircea Lazar is a researcher working at the intersection of control theory, optimization, and machine learning, with a focus on advancing intelligent and robust control systems for complex dynamical environments. His work spans risk-aware model predictive control (MPC), stochastic systems, and physics-guided neural networks, reflecting a commitment to bridging rigorous mathematical frameworks with practical engineering applications. Among his notable contributions is a pioneering approach to risk-aware MPC for stochastic systems with runtime temporal logic specifications — a significant departure from conventional methods that assume fixed, pre-specified constraints. By enabling dynamic assignment of temporal logic specifications during operation, Lazar's framework substantially broadens the applicability of formal control methods to real-world, unpredictable environments. This work has already attracted early citations since its 2024 publication. Lazar has also made contributions to physics-guided neural networks for inversion-based feedforward control, demonstrated in the context of hybrid stepper motors — systems critically important in robotics and industrial printing. This work illustrates his ability to integrate domain-specific physical knowledge with data-driven methods to improve control precision and efficiency without escalating manufacturing costs. Together, his research reflects a forward-looking vision for safe, adaptive, and computationally intelligent control systems.